Measurements and modelling of molecular iodine emissionstransport and photodestruction in the coastal region around Roscoff
R J Leigh1 S M Ball2 J Whitehead3 C Leblanc4 A J L Shillings5 A S Mahajan6 H Oetjen6 J D Lee7C E Jones7 J R Dorsey3 M Gallagher3 R L Jones5 J M C Plane6 P Potin4 and G McFiggans3
1Department of Physics and Astronomy University of Leicester Leicester UK2Department of Chemistry University of Leicester Leicester UK3School of Earth Atmospheric and Environmental Sciences University of Manchester Manchester UK4Station Biologique de Roscoff UPMC-CNRS UMR 7139 Roscoff France5Department of Chemistry University of Cambridge Cambridge UK6School of Chemistry University of Leeds Leeds UK7Department of Chemistry University of York York UK
Received 11 September 2009 ndash Published in Atmos Chem Phys Discuss 7 October 2009Revised 1 November 2010 ndash Accepted 5 November 2010 ndash Published 13 December 2010
Abstract Iodine emissions from the dominant six macroal-gal species in the coastal regions around Roscoff Francehave been modelled to support the Reactive Halogens inthe Marine Boundary Layer Experiment (RHaMBLe) under-taken in September 2006 A two-dimensional model is usedto explore the relationship between geographically resolvedregional emissions (based on maps of seaweed beds in thearea and seaweed I2 emission rates previously measured inthe laboratory) and in situ point and line measurements ofI2 performed respectively by a broadband cavity ringdownspectroscopy (BBCRDS) instrument sited on the shorelineand a long-path differential optical absorption spectroscopy(LP-DOAS) instrument sampling over an extended light pathto an off-shore island The modelled point and line I2 con-centrations compare quantitatively with BBCRDS and LP-DOAS measurements and provide a link between emis-sion fields and the different measurement geometries usedto quantify atmospheric I2 concentrations during RHaMBLeTotal I2 emissions over the 100 km2 region around Roscoffare calculated to be 17times1019 molecules per second duringthe lowest tides
During the night the model replicates I2 concentrations upto 50 pptv measured along the LP-DOAS instrumentrsquos lineof sight and predicts spikes of several hundred pptv in cer-tain conditions Point I2 concentrations up to 50 pptv arealso calculated at the measurement site in broad agreement
Correspondence toR J Leigh(rjleighleicesteracuk)
with the BBCRDS observations Daytime measured concen-trations of I2 at the site correlate with modelled productionand transport processes However substantial recycling ofthe photodissociated I2 is required for the model to quanti-tatively match measured concentrations This result corrob-orates previous modelling of iodine and NOx chemistry inthe semi-polluted marine boundary layer which proposed amechanism for recycling I2 via the formation transport andsubsequent reactions of the IONO2 reservoir compound
The methodology presented in this paper provides a toolfor linking spatially distinct measurements to inhomoge-neous and temporally varying emission fields
1 Introduction
Most techniques for measuring atmospheric compositionprovide a single point observation from which conclusionsare then sought over a wider region or scenario Othertechniques such as long-path differential optical absorptionspectroscopy (LP-DOAS) provide integrated measurementsalong a folded line of sight between a light source and retro-reflector The relationship between point and line data (and acampaignrsquos wider conclusions) can be understood by a rela-tively simple model linking the measurements to temporallyvarying and spatially inhomogeneous concentrations in theregion around the monitoring site Results are presentedhere from a model of molecular iodine emissions duringthe RHaMBLe campaign hosted at the Station Biologiquede Roscoff (SBR) in late summer 2006 (McFiggans et al
Published by Copernicus Publications on behalf of the European Geosciences Union
11824 R J Leigh et al Bridging spatial scales in measurements of I2
2010) This model is used to place novel measurements of I2from broadband cavity ring-down spectroscopy (BBCRDS)and LP-DOAS instruments into a regional context
Coastal emissions of reactive halogen gases merit inves-tigation owing to their linkages with perturbations of tro-pospheric radical chemistry aerosol particle nucleation andpossible climate impacts (von Glasow and Crutzen 2007McFiggans et al 2010) In coastal regions I2 has beenshown to be a significantly larger source of iodine atomsthan iodocarbons (McFiggans et al 2004) with elevated lo-calised concentrations of I2 measured at Mace Head Irelandpeaking around low tide (Saiz-Lopez and Plane 2004 Saiz-Lopez et al 2006) Mahajan et al(2009) also observed I2around low tides during RHaMBLe as discussed further inthis work
2 The model
The present model incorporates two horizontal spatial scalesand a temporal domain with an additional vertical com-ponent included in footprint modelling calculations Thehorizontal grid consists of 746 by 227 elements each of00005times00005 degrees extending fromminus42075 tominus3835degrees longitude and 486725 to 487855 degrees latitudeIn the Roscoff region this resolution corresponds to gridboxes of approximately 367 m longitudinally by 556 m lat-itudinally Bathymetry and macroalgal distribution informa-tion was mapped on to this model grid Tide and meteoro-logical data was applied to this spatial information at 1-minresolution from 5 to 28 September 2006 during the RHaM-BLe campaign
21 Seaweed speciation and site bathymetry
The Roscoff inter-tidal zone in front of the SBR extendsmore than five kilometers in length and about 1 km in widthMapping of the seaweed beds in the vicinity of Roscoffhas been attempted in two main studies One in the early1970s (Braud 1974) combined aerial photographs and insitu observations obtained from diving and field measure-ments The second study in the 1990s used both field andairborne spectrometers to map the seaweed and seagrass bedsnear Roscoff (Bajjouk et al 1996) The published mapsfrom these previous studies were used to construct a ded-icated map for this work which was further validated byfield observations from September 2006 to September 2009This seaweed map was then superposed with a bathymetrymap of the area provided by L Leveque from the Ser-vice Mer et Observation (Roscoff) The resulting mappeddistributions ofLaminaria digitata Laminaria hyperboreaLaminaria ochroleuca Saccharina latisima Fucus andAscophyllumare shown in Fig1 AscophyllumandFucusbeds are inherently mixed and are mapped together a con-
Fig 1 Bathymetry map and algal distributions used as inputs forthis modelling work The following key locations are marked themain measurement site(A) the LP-DOAS telescope(B) and theLP-DOAS retroreflector(C) Seaweed species are coded as followsL hyperboreandash purpleL digitatandash greenL ochroleucandash orangeSaccharina latissimandash yellowAscophyllumFucusndash red
stant mixing ratio of 6535 forAscophyllumFucuswas usedin modelling their emissions
The vertical zonation of seaweed species is very distincton rocky shores with each species often forming a belt at acertain elevation in the eulittoral zone (the area between thehighest and the lowest tides) and also in the subtidal zone(the area extending below the zero of the marine charts) Itis thought that the driving force of this zonation is a com-bination of biotic factors and the tolerance of the differentspecies to abiotic factors such as temperature light salinitydehydration and mechanical forces caused by wave action(Luning 1990) A typical kelp bed from the Roscoff regionis shown in Fig2
In the North Atlantic as exemplified in the study site infront of the SBR the eulittoral zone in sheltered habitatsis dominated both in coverage and biomass by brown algalspecies of the order of Fucales (fucoids) such asFucussppandAscophyllum nodosum In addition four species of theorder Laminariales (kelps) are distributed in distinct popula-tions forming beltsLaminaria digitataoccurs in the lowestpart of the eulittoral zone and in the upper subtidal zone withLaminaria hyperboreaextending from the upper subtidalzone to a limit of depth conditioned by the light penetration(about 20 m at Ile de Batz Table1) Laminaria ochroleucaappears in habitats protected from the dominant wind eithermixed withLaminaria hyperboreaandSaccharina lattisimaor in monospecific stands and mainly restricted to shallowwaters
R J Leigh et al Bridging spatial scales in measurements of I2 11825
2 R J Leigh Bridging spatial scales in measurements of I2
2 The Model
The present model incorporates two horizontal spatial scalesand a temporal domain with an additional vertical com-ponent included in footprint modelling calculations Thehorizontal grid consists of 746 by 227 elements each of00005times00005 degrees extending from -42075 to -3835degrees longitude and 486725 to 487855 degrees latitudeIn the Roscoff region this resolution corresponds to gridboxes of approximately 367 m longitudinally by 556 m lat-itudinally Bathymetry and macroalgal distribution informa-tion was mapped on to this model grid Tide and meteo-rological data was applied to this spatial information at 1-minute resolution from 5th to 28th September 2006 duringthe RHaMBLe campaign
21 Seaweed speciation and site bathymetry
The Roscoff inter-tidal zone in front of the SBR extendsmore than five kilometers in length and about 1 kilometerin width Mapping of the seaweed beds in the vicinity ofRoscoff has been attempted in two main studies One inthe early 1970s (Braud 1974) combined aerial photographsand in situ observations obtained from diving and field mea-surements The second study in the 1990s used both fieldand airborne spectrometers to map the seaweed and sea-grass beds near Roscoff (Bajjouk et al 1996) The pub-lished maps from these previous studies were used to con-struct a dedicated map for this work which was further vali-dated by field observations from September 2006 to Septem-ber 2009 This seaweed map was then superposed with abathymetry map of the area provided by L Leveque from theService Mer et Observation (Roscoff) The resulting mappeddistributions of Laminaria digitata Laminaria hyperboreaLaminaria ochroleuca Saccharina latisima Fucus andAscophyllum are shown in Figure 1 Ascophyllum and Fucusbeds are inherently mixed and are mapped together a con-stant mixing ratio of 6535 for AscophyllumFucus was usedin modelling their emissions
The vertical zonation of seaweed species is very distincton rocky shores with each species often forming a belt at acertain elevation in the eulittoral zone (the area between thehighest and the lowest tides) and also in the subtidal zone(the area extending below the zero of the marine charts) Itis thought that the driving force of this zonation is a com-bination of biotic factors and the tolerance of the differentspecies to abiotic factors such as temperature light salinitydehydration and mechanical forces caused by wave action(Luning 1990) A typical kelp bed from the Roscoff regionis shown in Figure 2
In the North Atlantic as exemplified in the study site infront of the SBR the eulittoral zone in sheltered habitatsis dominated both in coverage and biomass by brown algalspecies of the order of Fucales (fucoids) such as Fucus sppand Ascophyllum nodosum In addition four species of the
Fig 1 Bathymetry map and algal distributions used as inputs forthis modelling work The following key locations are marked themain measurement site (A) the LP-DOAS telescope (B) and theLP-DOAS retroreflector (C) Seaweed species are coded as followsL hyperborea - purple L digitata - green L ochroleuca - orangeSaccharina latissima - yellow AscophyllumFucus - red
Fig 2 The kelp bed at the Rocher du Loup at a low tide ofabout 05 m dominated by L digitata (kelp) and just above a beltof Himanthalia elongata (fucoid)
Fig 2 The kelp bed at the Rocher du Loup at a low tide ofabout 05 m dominated byL digitata (kelp) and just above a beltof Himanthalia elongata(fucoid)
The shallow inter-tidal zone at Roscoff results in the wa-ters closest to the shoreline being too shallow for Laminari-ales species So although a large horizontal surface area ofseaweed beds becomes exposed at low tide immediately infront of the SBR this mainly consists of fucoids (see Fig1)The distribution of seaweed species is rather patchy in theinter-tidal zone and is mainly dominated byFucus sppandAscophyllumbeds however there is also a small amount ofLdigitataSaccharina latissimaandL ochroleucain the chan-nel and tide pools between the site and the Ile de Batz islandto the north The south shore of Ile de Batz includes shel-tered shallow patchy habitats with sand and gravel which sur-round rocky areas covered by fucoids except where exposedto strong tide currentsLaminaria(mainly hyperborea) bedsextend to the north of Ile de Batz whereasL digitata flour-ishes in moderately exposed areas or at sites with strong wa-ter currents in the western part of the study site (Ile de Batzand islets west of Perharidy) and north-east from the Ile deBatzL digitataalso occurs in rockpools up to mid-tide leveland higher on wave-exposed coasts of the Ile de Batz
The average biomass densities in Table1 were obtainedfrom recent studies onAscophyllum nodosumat Roscoff(Gollety et al 2008) andL digitata (Gevaert et al 2008)and from the extensive long-term survey ofLaminarialespopulations by Ifremer (Arzel 1998) using the averagebiomass ofL digitata in September over the last ten yearsThe average volumic mass of each species was determinedexperimentally by filling a one-litre volume with seaweedthalli and determining the fresh weight of 5 replicates Thedepth limits of the various species at Roscoff were obtainedfrom previous mapping studies and from the Service Meret Observation in agreement with published data (Luning1990)
Fig 3 Emission rates assumed for the six species of macroalgae asa function of time following exposure to the atmosphere
22 Emission rates from exposed macroalgae
I2 emission rates for each species of macroalgae were esti-mated from the time since each model grid square was firstexposed to air by the changing tide and from the propen-sity of each seaweed species to emit For the former eachseaweed species was given a height attribute in the modelin order to account for the variable length and structure ofthe plants and therefore the variable water height at whichthe seaweed first breaks the water surface and becomes ex-posed to air These heights are shown in Table1 and re-sult in the I2 modelled emissions starting slightly in ad-vance of the surrounding sea bed itself becoming exposedSpecies specific I2 emission rates (in picomoles per minuteper gramme fresh weight) were parameterised from the lab-oratory study ofBall et al (2010) Figure3 shows the timedependent emission rates for each seaweed species used inthis work Fucusspecies andAscophyllumwere assumedto emit at a constant rate when exposed to air Emissionrates of theLaminaria species were assumed to decline af-ter first exposure to air with a common half life of 10 min-utes Laboratory studies have also found thatL digitatacansometimes resume bursts of strong I2 emission after havingbeen exposed to air for long periods (Dixneuf et al 2009Ball et al 2010) for simplicity here emission rates foreach species are assumed to become constant after 40 min-utes of continuous exposure Emissions from all speciescease immediately once the seaweed is re-covered by theincoming tide L ochroleuca for which measured emis-sion data were not available was assumed to emit at a rate
11826 R J Leigh et al Bridging spatial scales in measurements of I2
intermediate betweenL digitata and Saccharina latissimaThese emission rates were converted into emissions per m2
sea surface area using the assumptions shown in Table1 ofmass per m2 by species Emissions were assumed to mixinto an atmosphere layer of 15 cm depth providing a con-version into volume mixing ratio (VMR) This assumptionproduces peak VMRs immediately above the most strongly-emitting speciesL hyperboreaandL digitata of approxi-mately 15 ppbv (parts per billion by volume) immediately af-ter their first exposure to air consistent with the peak VMRsobserved byBall et al(2010) in their laboratory study
Actinic fluxes of solar radiation were measured using aMetcon spectral radiometer (Edwards and Monks 2003) andwere used to calculate the photolysis frequencies of a numberof trace gases including molecular iodine (jI2) ndash see Fig4The model temporal resolution was matched to the meteoro-logical dataset sampling of 1 min with all times within themodel expressed in Universal Time The model was appliedto data from 5 to 28 September 2006
23 Footprint analysis
Concentration footprints (as opposed to the more often usedflux footprints) were calculated for a range of wind veloc-ities and representative meteorological conditions using theanalytical approximation ofSchmid(1994) The model isa numerical solution to an analytical approximation of theadvection-diffusion equations The heterogeneity of the up-wind surface makes it rather difficult to draw firm conclu-sions about the exact form of the concentration footprintnevertheless the model is capable of providing sufficientlydetailed estimates for the purposes of this study
A surface roughness lengthz0 of 003 m was used wherethe fetch was across the inter-tidal zone For the water sur-faces encountered at high tidez0 was determined using therelationship described byZilitinkevich (1969)
z0= c1v
ulowast+
u2lowast
c2g(1)
whereulowast is the friction velocityg the acceleration due togravity v is the kinematic viscosity andc1 andc2 are coef-ficients with highly variable values estimated to be between00ndash048 (forc1) and frominfin to 811 (c2) In this studyintermediate values ofc1 = 01 andc2 = 320 were used Il-lustrations of footprint dimensions using this technique canbe found in Fig 13 ofMcFiggans et al(2010)
Using this technique footprints were calculated at fiveminute intervals throughout the campaign taking windspeedtide height and time of day as input parameters These inputparameters are shown in Fig4 Emission footprints werealso calculated atplusmn5 degrees either side of the wind direction(as measured at the SBR site) and the total modelled foot-print was taken to be the mean over all three of these footprintcalculations This averaging procedure aims to compensatefor temporal variability in wind direction within each 5 min
Fig 4 Tide I2 photolysis frequency and meteorological data usedto drive the model
time bin of the footprint calculation and for using wind datameasured at the SBR site to infer wind directionspeed overthe seaweed beds
Footprints calculated for each different tide height andwind strength were used in the model to characterise thetransport of I2 emissions to the site and into the LP DOAS in-strumentrsquos line of sight Following rotation appropriate to thewind direction the footprint for each time step was applied tothe emissions grid to estimate its contribution to the I2 con-centration observed by the BBCRDS and LP-DOAS instru-ments Footprints for the LP-DOAS were obtained throughintegration of all ldquopointrdquo footprints along the line of sight
3 Model results
The exposed regions of macroalgae were calculated for eachmodel time step and the resultant I2 emissions estimatedbased on the time since each grid cell had been exposed bythe tide and the seaweed species resident within the grid cellTwo example snap-shots of model footprints and emissionfields are shown in Figs5 and6 The curtain effect duringan ebb tide is shown in Fig5 as the initial exposure of theseaweed beds to air causes bursts of high I2 emissions espe-cially from the most potent emitters (red and green pixels)Emissions are lower and more uniform during the flow tide(see the dark blue pixels in Fig6)
Figure 7 shows the total regional emissions (panel 2)and the individual contributions from each seaweed species(panels 3ndash7) The greatest emissions correlated with the
R J Leigh et al Bridging spatial scales in measurements of I2 11827
Table 1 Derived bathymetry bands for each seaweed species included in this study with assumptions used to derive average fresh weightmass per m2 of coverage The final column details the height assumption for each species used to determine the tidal level at which seaweedbecomes exposed to the atmosphere
Species Minimum Maximum Average Average Average Averagedepth depth volumic mass volume biomass Height(m) (m) (kgFWm3) (m3m2) (kgFWm2) (m)
Fig 5 Model timestep from 1032 pm on 7 September 2006 dur-ing an ebb tide and one of the highest I2 concentrations predictedat site The wind speed and direction at this time were 565 msand 797 degrees respectively with the tide 085 m below the datumLP-DOAS and site footprints are shown in blue and grey shading re-spectively The modelled I2 emission fields are shown as red greenand dark blue pixels denoting emission rates of 1times1017 5times1016
and 25times1016molecules per grid square per second respectively
lowest low tides around 9ndash11 September and 23ndash25 Septem-ber when the largest area ofL digitata and particularlyL hyperboreabeds were uncovered The dominance of thesetwo Laminariaspecies in regional emissions is illustrated bythe the propensity of green and blue shading in panel 2 ofFig 7 these two species provide approximately 85 of to-tal regional emissions over the RHaMBLe campaign periodThe shapes of the emission profiles also change with the dif-ferent tidal spans The profiles are typically shorter and wideron days with the smallest tidal ranges around the middle ofthe campaign as seaweeds growing in shallow waters egfucusandascophyllumremain uncovered and thus contribut-ing their low level emissions throughout the majority of thetidal cycle In all cases the emission profiles are asymmetric
Fig 6 Model timestep from 548 pm on 14 September 2006 dur-ing a flow tide when the BBCRDS instrument measured signifi-cant concentrations of I2 The wind speed and direction at this timewere 19 ms and 309 degrees respectively with the tide 183 metresabove the datum The modelled I2 emission fields are shown on thesame scale as Fig5with dark blue purple and black pixels denotingemission rates of 25times1016 125times1016 andle 50times1015moleculesper grid square per second respectively The reduction in emissionssince first exposure can be seen with respect to Fig5
being biased towards greater emissions when the seaweedsare first uncovered by the retreating tide the initial burst ofemissions following first exposure is evident in the contribu-tions ofL ochroleucaand particularlySaccharina latissima(which grows in habitats spanning a narrow depth range)
4 I2 measurements during RHaMBLe
41 BBCRDS measurements
A broadband cavity ringdown spectrometer was deployedfrom a shipping container sited on the jetty in front of theSBR adjacent to the containers housing the campaignrsquos other
11828 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 7 Tide height (panel 1) and calculated regional emissions of I2 from 5 to 28 September 2006 (panel 2) The emission contributions aredifferentiated according to seaweed species by colour in panel 2 and are plotted separately in panels 3ndash7
in situ instruments (McFiggans et al 2010) Broadbandcavity ringdown spectroscopy (BBCRDS) uses light from apulsed broadband laser to measure the absorption spectrumof samples contained within a high finesse optical cavity(Bitter et al 2005 Ball and Jones 2003 2009) In this casethe BBCRDS instrument was configured to detect molecu-lar iodine using several of the I2 moleculersquos Blarr X absorp-tion bands in the wavelength range 560ndash570 nm Other at-mospheric gases (H2O NO2 and the oxygen dimer O4) alsoabsorb at these wavelengths and thus contribute to the mea-sured BBCRDS spectra
The BBCRDS system used for this study is based on aninstrument previously used to measure I2 at Mace Head (Ire-land) during the 2002 NAMBLEX campaign (Saiz-Lopezet al 2006 Heard et al 2006) as described in detail byBitteret al (2005) In the intervening years the instrumentrsquos per-formance has been enhanced significantly by upgrading sev-eral key components notably a new laser system that yieldspulsed broadband light with a factor of two wider bandwidth
at green wavelengths a new clocked CCD camera and im-proved analysis softwarespectral fitting routines A broad-band dye laser pumped by a 532 nm NdYAG laser (Sirah Co-bra and Surelight I-20 20 Hz repetition rate) generated lightpulses with an approximately Gaussian emission spectrumcentred at 563 nm (FWHM = 52 nm) This light was directedinto a 187 cm long ringdown cavity formed by two highlyreflective mirrors (Los Gatos peak reflectivity = 99993 at570 nm) Light exiting the ringdown cavity was collected andconveyed through a 100 microm core diameter fibre optic cable toan imaging spectrograph (Chromex 250is) where it was dis-persed in wavelength and imaged onto a clocked CCD cam-era (XCam CCDRem2) The time evolution of individualringdown events was recorded simultaneously at 512 differ-ent wavelengths one for each pixel row of the detector andlight from 50 ringdown events was integrated on the CCDcamera before storing the data to a computer Wavelengthresolved ringdown times were produced by fitting the ring-down decay in each pixel row (j = 1 to 512) The samplersquos
R J Leigh et al Bridging spatial scales in measurements of I2 11829
absorption spectrum was then calculated from sets of ring-down times measured when the cavity contained the sampleτ(λj ) and when flushed with dry nitrogenτ0(λj )
α(λj ) =RL
c
(1
τ(λj )minus
1
τ0(λj )
)=
sumn
αn(λj )+αcon(λj ) (2)
wherec is the speed of light RL is the fraction of the cavitythat is occupied by absorbing speciesαn(λj ) is the wave-length dependent absorption coefficient of the nth molecu-lar absorber andαcon(λj ) is the absorption coefficient dueto all other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16 September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Eq2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken fromVandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported bySaiz-Lopez et al(2004) (see alsoBall et al (2010)) The toppanel of Fig8 shows an example BBCRDS spectrum ob-tained during the campaign where the central and lower pan-els show respectively the I2 and NO2 contributions to themeasured absorption overlaid by their fitted reference spectrafrom the DOAS fitting routine During the RHaMBLe cam-paign the precision of the spectral retrievals was typically10 pptv (parts per trillion by volume) for I2 and 02 ppbv forNO2 (1σ uncertainty 304 s averaging time) Although notthe principal target of this deployment co-retrieval of theNO2 concentrations served as an important quality assuranceparameter with which to monitor the BBCRDS instrumentrsquosperformance Throughout the campaign the NO2 concen-trations measured by BBCRDS were in excellent quantita-tive agreement with NO2 measurements made by the Uni-versity of Yorkrsquos NOxy chemiluminescence instrument as
R J Leigh Bridging spatial scales in measurements of I2 7
α(λj) =RL
c
(1
τ(λj)minus 1τ0(λj)
)=sum
n
αn(λj) + αcon(λj) (2)
where c is the speed of light RL is the fraction of the cav-ity that is occupied by absorbing species αn(λj) is the wave-length dependent absorption coefficient of the nth molecularabsorber and αcon(λj) is the absorption coefficient due toall other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16th September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Equa-tion 2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2
were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken from Vandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported by Saiz-Lopez et al (2004) (see also Ball et al (2010)) The toppanel of Figure 8 shows an example BBCRDS spectrumobtained during the campaign where the central and lowerpanels show respectively the I2 and NO2 contributions tothe measured absorption overlaid by their fitted referencespectra from the DOAS fitting routine During the RHaM-BLe campaign the precision of the spectral retrievals wastypically 10 pptv for I2 and 02 ppbv for NO2 (1σ uncer-tainty 304 s averaging time) Although not the principaltarget of this deployment co-retrieval of the NO2 concentra-tions served as an important quality assurance parameter withwhich to monitor the BBCRDS instrumentrsquos performanceThroughout the campaign the NO2 concentrations measuredby BBCRDS were in excellent quantitative agreement withNO2 measurements made by the University of Yorkrsquos NOxy
chemiluminescence instrument as described in McFigganset al (2010) The NO2 amounts are also a valuable indicator
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14th September 2006 The upper panel shows the measuredspectrum (blue) after subtraction of the absorptions due to watervapour O4 and a second order polynomial accounting for the un-structured absorption contributions The red line shows a DOAS fitto the spectrumrsquos differential structure and the residual spectrumis shown in green The measured (blue) and fitted (red) absorptioncontributions due to I2 and NO2 are shown in the middle and lowerpanels respectively
of the possible extent of I2 recycling via IONO2 chemistryin the semi-polluted environment around Roscoff and so theNO2 field observations from both instruments are shown to-gether in the figures illustrating the measured and modelledI2 concentrations (see Figs 10 12 16 18) The generallygood agreement between the BBCRDS and chemilumines-cence measurements across a wide range of rapidly varyingNO2 concentrations is exemplified by the data from 14th-15th September shown in the bottom panel of Figure 10the gradient of a correlation plot of the NO2 concentrationsrecorded by the two instruments was 098 plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathDifferential Optical Absorption Spectroscopy (LP-DOAS)technique (Plane and Saiz-Lopez 2006) was used to mea-
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14 September 2006 The upper panel shows the measured spec-trum (blue) after subtraction of the absorptions due to water vapourO4 and a second order polynomial accounting for the unstructuredabsorption contributions The red line shows a DOAS fit to the spec-trumrsquos differential structure and the residual spectrum is shown ingreen The measured (blue) and fitted (red) absorption contribu-tions due to I2 and NO2 are shown in the middle and lower panelsrespectively
described inMcFiggans et al(2010) The NO2 amountsare also a valuable indicator of the possible extent of I2recycling via IONO2 chemistry in the semi-polluted envi-ronment around Roscoff and so the NO2 field observationsfrom both instruments are shown together in the figures il-lustrating the measured and modelled I2 concentrations (seeFigs10 12 16 and18) The generally good agreement be-tween the BBCRDS and chemiluminescence measurementsacross a wide range of rapidly varying NO2 concentrations isexemplified by the data from 14ndash15 September shown in thebottom panel of Fig 10 the gradient of a correlation plot ofthe NO2 concentrations recorded by the two instruments was098plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathdifferential optical absorption spectroscopy (LP-DOAS)
11830 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 9 Complete timeseries of model output from 5 to 28 Septem-ber 2006 From top down Regional emissions as calculated bythe model I2 concentrations at the measurement site based on foot-print analysis calculated with photolytic destruction and no recy-cling processes likewise but with recycling at 95 The lowesttwo panels show calculated I2 concentrations along the LP-DOASline of sight with photolytic destruction and no recycling processesand with recycling at 95
technique (Plane and Saiz-Lopez 2006) was used to mea-sure the concentrations of I2 OIO IO and NO3 The absorp-tion path extended 335 km from the SBR (48728 latitudeminus3988 longitude) to a small outcrop on the south west shoreof the Ile de Batz (4874 latitudeminus4036 longitude) wherea retroreflector array was placed to fold the optical path ndash seealso Fig1 The total optical path length was thus 67 km withthe beam 7 to 12 m above the mean sea level Full detailsof the DOAS instrument can be found elsewhere (Mahajanet al 2009 Saiz-Lopez and Plane 2004)
Briefly spectra were recorded with 025 nm resolution be-fore being converted into differential optical density spec-tra The contributions of individual absorbing species to the
Fig 10 Modelled and measured data from 13 to 15 September2006 The top panel shows BBCRDS data (red points and errorbars) with modelled concentrations of I2 at the site assuming 95recycling of I2 photolysed during the daytime (orange line) Themiddle panel shows LP-DOAS data (dark blue points and error bars)with modelled concentrations of I2 in the LP-DOAS light path as-suming 95 I2 recycling (blue line) The grey areas in the uppertwo plots indicate the range of modelled I2 values using recyclingassumptions fromR = 90 to 98 The photolysis frequency of I2is indicated by the green line in the upper two plots and tide by theblack line The bottom panel shows NO2 measured by the NOxychemiluminescence (black) and BBCRDS instruments (red pointsand error bars)
measured spectrum were determined by simultaneous fittingof their molecular absorption cross sections using singularvalue decomposition (Plane and Saiz-Lopez 2006) Aver-aged I2 concentrations along the line of sight were retrievedin the 535minus575 nm window on a number of days and nightsusing the I2 absorption cross sections ofSaiz-Lopez et al(2004) The full data set from the LP-DOAS instrumentis presented inMahajan et al(2009) andMcFiggans et al(2010)
For the present work footprints for the LP-DOAS instru-ment were calculated using the same footprint model (as-suming an 8 m height for the LP-DOAS light beam) withmodelled I2 amounts averaged for the footprints along theline of sight In this way the model provides a path lengthaveraged measurement of I2 along the LP-DOAS light pathwhich sampled emissions from a significant proportion of the
R J Leigh et al Bridging spatial scales in measurements of I2 11831
channel between the SBR and the Ile de Batz (see for exam-ple the footprints in Figs5 and6)
5 Calculation of total emissions and modelled I2 at thesite and along the LP DOAS line of sight
Spatially resolved emissions were calculated for each gridsquare within the model for each one minute time step Siteand LP-DOAS footprints were calculated taking account ofwind speed wind direction and tidal height The time takenfor I2 emissions to travel from their source within the foot-print to the detector was estimated from the windspeed mea-sured at the site In addition to dilution of the emitted I2(accounted for in the footprint) the model also consideredthe photolytic destruction of I2 during its transit to the mea-surement location using equation3
[I2](t)=[I2](0)ej I2middott (3)
where [I2](t) is the volume mixing ratio of I2 at time t and[I2](0) is the volume mixing ratio of I2 at time 0 directlyabove the emission source jI2 is the photolysis frequencyof I2 as measured by a spectral radiometer
Although no chemical modelling was attempted along thelines ofMahajan et al(2009) a simple recycling parameterR was also included in this work to mimic the effects of I2recycling via IONO2 chemistry downstream of the emissionsource This recycling was achieved through modification ofthe I2 photolytic destruction process to
[I2](t)=[I2](0)ej I2middot(1minusR)middott (4)
The recycling parameterR effectively permits a propor-tion of photodissociated I2 to be instantly reformed in ourmodel In order to test this approach the chemical model pre-sented previously inMahajan et al(2009) was re-run in thisstudy to examine the decay of I2 concentrations downwindof an emission source The chemical model was run twiceonce in the presence of NOx (for an assumed baseline NO2concentration of 2 ppbv) and again in the absence of NOxbut for reduced I2 photolysis rates (whilst keeping the pho-tolysis rates of other photolabile species unchanged) Goodqualitative and reasonable quantitative agreement was foundbetween these two scenarios when the I2 photolysis rate wasdecreased to 10 of its typical daytime value providing avalidation for our simplistic approach of reducing the effec-tive I2 photolysis rates to mimic the effect of I2 recycling viaIONO2 formation in this semi-polluted atmosphere For thepresent modelling study the recycling parameter was set to adefault value ofR = 095 ie 95 of the I2 that is photolysedis reformed by subsequent chemistry This is higher than theR = 090 suggested by theMahajan et al(2009) chemicalmodelling but tended to produce the best agreement withthe observational data For comparison modelled I2 concen-trations were also calculated for smaller (R = 090) and larger
(R=098) recycling efficiencies appropriate for respectivelyNO2 concentrations at and above the NO2 = 2 ppbv baselinecase of theMahajan et al(2009) model These results areadditionally shown in Fig 10 and the following other resultsfigures
Figure 9 shows total calculated regional emissions themodelled I2 mixing ratios in air advected to the measurementsite (BBCRDS) and the mean I2 mixing ratio in the air sam-pled along the LP-DOAS line of sight Modelled emissionsin the middle and bottom panels assume 95 recycling ofphotolysed I2 during the daytime Fig9 considers two I2 lossscenarios (i) dilution and irreversible photolytic loss accord-ing to Eq (3) and (ii) dilution and photolytic loss less a re-cycling assumption atR = 095 as per Eq (4) The modelledand measured I2 concentrations are compared in the follow-ing section
6 Comparison of modelled and measured I2
Owing to the challenges inherent in operating the BBCRDSand LP-DOAS instruments in the field and the requirementsfor measurements of other species to be taken by the sameinstrumentation I2 observations are unfortunately not avail-able throughout the campaign Results from three intensivemeasurement periods from each instrument are detailed be-low
61 Comparisons during the night
During the night in the absence of photolytic destructionI2 can be considered as a passive tracer to establish the linkbetween emissions and measurements Agreement betweenthe model and BBCRDS and LP-DOAS measurements atnight indicates acceptable model parameterisation of emis-sion rates seaweed spatial distribution meteorology and di-lution and dispersion within the emission footprint This cantherefore be used as a baseline for investigations during theday when additionally photolytic destruction and chemicalrecycling of I2 become important
Figure10 illustrates night-time measurements and modeldata from 13 to 15 September while Fig11 shows the cor-responding sources of I2 (differentiated by seaweed species)modelled at the measurement site and along the LP-DOASlight path In Fig10 the diurnal cycle is indicated by thephotolysis frequency of I2 (green line) with tide heightsshown in black During the period described in Fig10 therewere night-time low tides shortly after midnight on 14 and15 September 2006 when respectively both the LP-DOASand BBCRDS observed I2 substantially above their detec-tion limits and the model predicted peak I2 concentrations of50ndash100 pptv
Three modelled I2 datasets are presented in Fig10 (andother following data figures) covering the range of I2 re-cycling assumptions discussed in the previous section I2
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
11824 R J Leigh et al Bridging spatial scales in measurements of I2
2010) This model is used to place novel measurements of I2from broadband cavity ring-down spectroscopy (BBCRDS)and LP-DOAS instruments into a regional context
Coastal emissions of reactive halogen gases merit inves-tigation owing to their linkages with perturbations of tro-pospheric radical chemistry aerosol particle nucleation andpossible climate impacts (von Glasow and Crutzen 2007McFiggans et al 2010) In coastal regions I2 has beenshown to be a significantly larger source of iodine atomsthan iodocarbons (McFiggans et al 2004) with elevated lo-calised concentrations of I2 measured at Mace Head Irelandpeaking around low tide (Saiz-Lopez and Plane 2004 Saiz-Lopez et al 2006) Mahajan et al(2009) also observed I2around low tides during RHaMBLe as discussed further inthis work
2 The model
The present model incorporates two horizontal spatial scalesand a temporal domain with an additional vertical com-ponent included in footprint modelling calculations Thehorizontal grid consists of 746 by 227 elements each of00005times00005 degrees extending fromminus42075 tominus3835degrees longitude and 486725 to 487855 degrees latitudeIn the Roscoff region this resolution corresponds to gridboxes of approximately 367 m longitudinally by 556 m lat-itudinally Bathymetry and macroalgal distribution informa-tion was mapped on to this model grid Tide and meteoro-logical data was applied to this spatial information at 1-minresolution from 5 to 28 September 2006 during the RHaM-BLe campaign
21 Seaweed speciation and site bathymetry
The Roscoff inter-tidal zone in front of the SBR extendsmore than five kilometers in length and about 1 km in widthMapping of the seaweed beds in the vicinity of Roscoffhas been attempted in two main studies One in the early1970s (Braud 1974) combined aerial photographs and insitu observations obtained from diving and field measure-ments The second study in the 1990s used both field andairborne spectrometers to map the seaweed and seagrass bedsnear Roscoff (Bajjouk et al 1996) The published mapsfrom these previous studies were used to construct a ded-icated map for this work which was further validated byfield observations from September 2006 to September 2009This seaweed map was then superposed with a bathymetrymap of the area provided by L Leveque from the Ser-vice Mer et Observation (Roscoff) The resulting mappeddistributions ofLaminaria digitata Laminaria hyperboreaLaminaria ochroleuca Saccharina latisima Fucus andAscophyllumare shown in Fig1 AscophyllumandFucusbeds are inherently mixed and are mapped together a con-
Fig 1 Bathymetry map and algal distributions used as inputs forthis modelling work The following key locations are marked themain measurement site(A) the LP-DOAS telescope(B) and theLP-DOAS retroreflector(C) Seaweed species are coded as followsL hyperboreandash purpleL digitatandash greenL ochroleucandash orangeSaccharina latissimandash yellowAscophyllumFucusndash red
stant mixing ratio of 6535 forAscophyllumFucuswas usedin modelling their emissions
The vertical zonation of seaweed species is very distincton rocky shores with each species often forming a belt at acertain elevation in the eulittoral zone (the area between thehighest and the lowest tides) and also in the subtidal zone(the area extending below the zero of the marine charts) Itis thought that the driving force of this zonation is a com-bination of biotic factors and the tolerance of the differentspecies to abiotic factors such as temperature light salinitydehydration and mechanical forces caused by wave action(Luning 1990) A typical kelp bed from the Roscoff regionis shown in Fig2
In the North Atlantic as exemplified in the study site infront of the SBR the eulittoral zone in sheltered habitatsis dominated both in coverage and biomass by brown algalspecies of the order of Fucales (fucoids) such asFucussppandAscophyllum nodosum In addition four species of theorder Laminariales (kelps) are distributed in distinct popula-tions forming beltsLaminaria digitataoccurs in the lowestpart of the eulittoral zone and in the upper subtidal zone withLaminaria hyperboreaextending from the upper subtidalzone to a limit of depth conditioned by the light penetration(about 20 m at Ile de Batz Table1) Laminaria ochroleucaappears in habitats protected from the dominant wind eithermixed withLaminaria hyperboreaandSaccharina lattisimaor in monospecific stands and mainly restricted to shallowwaters
R J Leigh et al Bridging spatial scales in measurements of I2 11825
2 R J Leigh Bridging spatial scales in measurements of I2
2 The Model
The present model incorporates two horizontal spatial scalesand a temporal domain with an additional vertical com-ponent included in footprint modelling calculations Thehorizontal grid consists of 746 by 227 elements each of00005times00005 degrees extending from -42075 to -3835degrees longitude and 486725 to 487855 degrees latitudeIn the Roscoff region this resolution corresponds to gridboxes of approximately 367 m longitudinally by 556 m lat-itudinally Bathymetry and macroalgal distribution informa-tion was mapped on to this model grid Tide and meteo-rological data was applied to this spatial information at 1-minute resolution from 5th to 28th September 2006 duringthe RHaMBLe campaign
21 Seaweed speciation and site bathymetry
The Roscoff inter-tidal zone in front of the SBR extendsmore than five kilometers in length and about 1 kilometerin width Mapping of the seaweed beds in the vicinity ofRoscoff has been attempted in two main studies One inthe early 1970s (Braud 1974) combined aerial photographsand in situ observations obtained from diving and field mea-surements The second study in the 1990s used both fieldand airborne spectrometers to map the seaweed and sea-grass beds near Roscoff (Bajjouk et al 1996) The pub-lished maps from these previous studies were used to con-struct a dedicated map for this work which was further vali-dated by field observations from September 2006 to Septem-ber 2009 This seaweed map was then superposed with abathymetry map of the area provided by L Leveque from theService Mer et Observation (Roscoff) The resulting mappeddistributions of Laminaria digitata Laminaria hyperboreaLaminaria ochroleuca Saccharina latisima Fucus andAscophyllum are shown in Figure 1 Ascophyllum and Fucusbeds are inherently mixed and are mapped together a con-stant mixing ratio of 6535 for AscophyllumFucus was usedin modelling their emissions
The vertical zonation of seaweed species is very distincton rocky shores with each species often forming a belt at acertain elevation in the eulittoral zone (the area between thehighest and the lowest tides) and also in the subtidal zone(the area extending below the zero of the marine charts) Itis thought that the driving force of this zonation is a com-bination of biotic factors and the tolerance of the differentspecies to abiotic factors such as temperature light salinitydehydration and mechanical forces caused by wave action(Luning 1990) A typical kelp bed from the Roscoff regionis shown in Figure 2
In the North Atlantic as exemplified in the study site infront of the SBR the eulittoral zone in sheltered habitatsis dominated both in coverage and biomass by brown algalspecies of the order of Fucales (fucoids) such as Fucus sppand Ascophyllum nodosum In addition four species of the
Fig 1 Bathymetry map and algal distributions used as inputs forthis modelling work The following key locations are marked themain measurement site (A) the LP-DOAS telescope (B) and theLP-DOAS retroreflector (C) Seaweed species are coded as followsL hyperborea - purple L digitata - green L ochroleuca - orangeSaccharina latissima - yellow AscophyllumFucus - red
Fig 2 The kelp bed at the Rocher du Loup at a low tide ofabout 05 m dominated by L digitata (kelp) and just above a beltof Himanthalia elongata (fucoid)
Fig 2 The kelp bed at the Rocher du Loup at a low tide ofabout 05 m dominated byL digitata (kelp) and just above a beltof Himanthalia elongata(fucoid)
The shallow inter-tidal zone at Roscoff results in the wa-ters closest to the shoreline being too shallow for Laminari-ales species So although a large horizontal surface area ofseaweed beds becomes exposed at low tide immediately infront of the SBR this mainly consists of fucoids (see Fig1)The distribution of seaweed species is rather patchy in theinter-tidal zone and is mainly dominated byFucus sppandAscophyllumbeds however there is also a small amount ofLdigitataSaccharina latissimaandL ochroleucain the chan-nel and tide pools between the site and the Ile de Batz islandto the north The south shore of Ile de Batz includes shel-tered shallow patchy habitats with sand and gravel which sur-round rocky areas covered by fucoids except where exposedto strong tide currentsLaminaria(mainly hyperborea) bedsextend to the north of Ile de Batz whereasL digitata flour-ishes in moderately exposed areas or at sites with strong wa-ter currents in the western part of the study site (Ile de Batzand islets west of Perharidy) and north-east from the Ile deBatzL digitataalso occurs in rockpools up to mid-tide leveland higher on wave-exposed coasts of the Ile de Batz
The average biomass densities in Table1 were obtainedfrom recent studies onAscophyllum nodosumat Roscoff(Gollety et al 2008) andL digitata (Gevaert et al 2008)and from the extensive long-term survey ofLaminarialespopulations by Ifremer (Arzel 1998) using the averagebiomass ofL digitata in September over the last ten yearsThe average volumic mass of each species was determinedexperimentally by filling a one-litre volume with seaweedthalli and determining the fresh weight of 5 replicates Thedepth limits of the various species at Roscoff were obtainedfrom previous mapping studies and from the Service Meret Observation in agreement with published data (Luning1990)
Fig 3 Emission rates assumed for the six species of macroalgae asa function of time following exposure to the atmosphere
22 Emission rates from exposed macroalgae
I2 emission rates for each species of macroalgae were esti-mated from the time since each model grid square was firstexposed to air by the changing tide and from the propen-sity of each seaweed species to emit For the former eachseaweed species was given a height attribute in the modelin order to account for the variable length and structure ofthe plants and therefore the variable water height at whichthe seaweed first breaks the water surface and becomes ex-posed to air These heights are shown in Table1 and re-sult in the I2 modelled emissions starting slightly in ad-vance of the surrounding sea bed itself becoming exposedSpecies specific I2 emission rates (in picomoles per minuteper gramme fresh weight) were parameterised from the lab-oratory study ofBall et al (2010) Figure3 shows the timedependent emission rates for each seaweed species used inthis work Fucusspecies andAscophyllumwere assumedto emit at a constant rate when exposed to air Emissionrates of theLaminaria species were assumed to decline af-ter first exposure to air with a common half life of 10 min-utes Laboratory studies have also found thatL digitatacansometimes resume bursts of strong I2 emission after havingbeen exposed to air for long periods (Dixneuf et al 2009Ball et al 2010) for simplicity here emission rates foreach species are assumed to become constant after 40 min-utes of continuous exposure Emissions from all speciescease immediately once the seaweed is re-covered by theincoming tide L ochroleuca for which measured emis-sion data were not available was assumed to emit at a rate
11826 R J Leigh et al Bridging spatial scales in measurements of I2
intermediate betweenL digitata and Saccharina latissimaThese emission rates were converted into emissions per m2
sea surface area using the assumptions shown in Table1 ofmass per m2 by species Emissions were assumed to mixinto an atmosphere layer of 15 cm depth providing a con-version into volume mixing ratio (VMR) This assumptionproduces peak VMRs immediately above the most strongly-emitting speciesL hyperboreaandL digitata of approxi-mately 15 ppbv (parts per billion by volume) immediately af-ter their first exposure to air consistent with the peak VMRsobserved byBall et al(2010) in their laboratory study
Actinic fluxes of solar radiation were measured using aMetcon spectral radiometer (Edwards and Monks 2003) andwere used to calculate the photolysis frequencies of a numberof trace gases including molecular iodine (jI2) ndash see Fig4The model temporal resolution was matched to the meteoro-logical dataset sampling of 1 min with all times within themodel expressed in Universal Time The model was appliedto data from 5 to 28 September 2006
23 Footprint analysis
Concentration footprints (as opposed to the more often usedflux footprints) were calculated for a range of wind veloc-ities and representative meteorological conditions using theanalytical approximation ofSchmid(1994) The model isa numerical solution to an analytical approximation of theadvection-diffusion equations The heterogeneity of the up-wind surface makes it rather difficult to draw firm conclu-sions about the exact form of the concentration footprintnevertheless the model is capable of providing sufficientlydetailed estimates for the purposes of this study
A surface roughness lengthz0 of 003 m was used wherethe fetch was across the inter-tidal zone For the water sur-faces encountered at high tidez0 was determined using therelationship described byZilitinkevich (1969)
z0= c1v
ulowast+
u2lowast
c2g(1)
whereulowast is the friction velocityg the acceleration due togravity v is the kinematic viscosity andc1 andc2 are coef-ficients with highly variable values estimated to be between00ndash048 (forc1) and frominfin to 811 (c2) In this studyintermediate values ofc1 = 01 andc2 = 320 were used Il-lustrations of footprint dimensions using this technique canbe found in Fig 13 ofMcFiggans et al(2010)
Using this technique footprints were calculated at fiveminute intervals throughout the campaign taking windspeedtide height and time of day as input parameters These inputparameters are shown in Fig4 Emission footprints werealso calculated atplusmn5 degrees either side of the wind direction(as measured at the SBR site) and the total modelled foot-print was taken to be the mean over all three of these footprintcalculations This averaging procedure aims to compensatefor temporal variability in wind direction within each 5 min
Fig 4 Tide I2 photolysis frequency and meteorological data usedto drive the model
time bin of the footprint calculation and for using wind datameasured at the SBR site to infer wind directionspeed overthe seaweed beds
Footprints calculated for each different tide height andwind strength were used in the model to characterise thetransport of I2 emissions to the site and into the LP DOAS in-strumentrsquos line of sight Following rotation appropriate to thewind direction the footprint for each time step was applied tothe emissions grid to estimate its contribution to the I2 con-centration observed by the BBCRDS and LP-DOAS instru-ments Footprints for the LP-DOAS were obtained throughintegration of all ldquopointrdquo footprints along the line of sight
3 Model results
The exposed regions of macroalgae were calculated for eachmodel time step and the resultant I2 emissions estimatedbased on the time since each grid cell had been exposed bythe tide and the seaweed species resident within the grid cellTwo example snap-shots of model footprints and emissionfields are shown in Figs5 and6 The curtain effect duringan ebb tide is shown in Fig5 as the initial exposure of theseaweed beds to air causes bursts of high I2 emissions espe-cially from the most potent emitters (red and green pixels)Emissions are lower and more uniform during the flow tide(see the dark blue pixels in Fig6)
Figure 7 shows the total regional emissions (panel 2)and the individual contributions from each seaweed species(panels 3ndash7) The greatest emissions correlated with the
R J Leigh et al Bridging spatial scales in measurements of I2 11827
Table 1 Derived bathymetry bands for each seaweed species included in this study with assumptions used to derive average fresh weightmass per m2 of coverage The final column details the height assumption for each species used to determine the tidal level at which seaweedbecomes exposed to the atmosphere
Species Minimum Maximum Average Average Average Averagedepth depth volumic mass volume biomass Height(m) (m) (kgFWm3) (m3m2) (kgFWm2) (m)
Fig 5 Model timestep from 1032 pm on 7 September 2006 dur-ing an ebb tide and one of the highest I2 concentrations predictedat site The wind speed and direction at this time were 565 msand 797 degrees respectively with the tide 085 m below the datumLP-DOAS and site footprints are shown in blue and grey shading re-spectively The modelled I2 emission fields are shown as red greenand dark blue pixels denoting emission rates of 1times1017 5times1016
and 25times1016molecules per grid square per second respectively
lowest low tides around 9ndash11 September and 23ndash25 Septem-ber when the largest area ofL digitata and particularlyL hyperboreabeds were uncovered The dominance of thesetwo Laminariaspecies in regional emissions is illustrated bythe the propensity of green and blue shading in panel 2 ofFig 7 these two species provide approximately 85 of to-tal regional emissions over the RHaMBLe campaign periodThe shapes of the emission profiles also change with the dif-ferent tidal spans The profiles are typically shorter and wideron days with the smallest tidal ranges around the middle ofthe campaign as seaweeds growing in shallow waters egfucusandascophyllumremain uncovered and thus contribut-ing their low level emissions throughout the majority of thetidal cycle In all cases the emission profiles are asymmetric
Fig 6 Model timestep from 548 pm on 14 September 2006 dur-ing a flow tide when the BBCRDS instrument measured signifi-cant concentrations of I2 The wind speed and direction at this timewere 19 ms and 309 degrees respectively with the tide 183 metresabove the datum The modelled I2 emission fields are shown on thesame scale as Fig5with dark blue purple and black pixels denotingemission rates of 25times1016 125times1016 andle 50times1015moleculesper grid square per second respectively The reduction in emissionssince first exposure can be seen with respect to Fig5
being biased towards greater emissions when the seaweedsare first uncovered by the retreating tide the initial burst ofemissions following first exposure is evident in the contribu-tions ofL ochroleucaand particularlySaccharina latissima(which grows in habitats spanning a narrow depth range)
4 I2 measurements during RHaMBLe
41 BBCRDS measurements
A broadband cavity ringdown spectrometer was deployedfrom a shipping container sited on the jetty in front of theSBR adjacent to the containers housing the campaignrsquos other
11828 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 7 Tide height (panel 1) and calculated regional emissions of I2 from 5 to 28 September 2006 (panel 2) The emission contributions aredifferentiated according to seaweed species by colour in panel 2 and are plotted separately in panels 3ndash7
in situ instruments (McFiggans et al 2010) Broadbandcavity ringdown spectroscopy (BBCRDS) uses light from apulsed broadband laser to measure the absorption spectrumof samples contained within a high finesse optical cavity(Bitter et al 2005 Ball and Jones 2003 2009) In this casethe BBCRDS instrument was configured to detect molecu-lar iodine using several of the I2 moleculersquos Blarr X absorp-tion bands in the wavelength range 560ndash570 nm Other at-mospheric gases (H2O NO2 and the oxygen dimer O4) alsoabsorb at these wavelengths and thus contribute to the mea-sured BBCRDS spectra
The BBCRDS system used for this study is based on aninstrument previously used to measure I2 at Mace Head (Ire-land) during the 2002 NAMBLEX campaign (Saiz-Lopezet al 2006 Heard et al 2006) as described in detail byBitteret al (2005) In the intervening years the instrumentrsquos per-formance has been enhanced significantly by upgrading sev-eral key components notably a new laser system that yieldspulsed broadband light with a factor of two wider bandwidth
at green wavelengths a new clocked CCD camera and im-proved analysis softwarespectral fitting routines A broad-band dye laser pumped by a 532 nm NdYAG laser (Sirah Co-bra and Surelight I-20 20 Hz repetition rate) generated lightpulses with an approximately Gaussian emission spectrumcentred at 563 nm (FWHM = 52 nm) This light was directedinto a 187 cm long ringdown cavity formed by two highlyreflective mirrors (Los Gatos peak reflectivity = 99993 at570 nm) Light exiting the ringdown cavity was collected andconveyed through a 100 microm core diameter fibre optic cable toan imaging spectrograph (Chromex 250is) where it was dis-persed in wavelength and imaged onto a clocked CCD cam-era (XCam CCDRem2) The time evolution of individualringdown events was recorded simultaneously at 512 differ-ent wavelengths one for each pixel row of the detector andlight from 50 ringdown events was integrated on the CCDcamera before storing the data to a computer Wavelengthresolved ringdown times were produced by fitting the ring-down decay in each pixel row (j = 1 to 512) The samplersquos
R J Leigh et al Bridging spatial scales in measurements of I2 11829
absorption spectrum was then calculated from sets of ring-down times measured when the cavity contained the sampleτ(λj ) and when flushed with dry nitrogenτ0(λj )
α(λj ) =RL
c
(1
τ(λj )minus
1
τ0(λj )
)=
sumn
αn(λj )+αcon(λj ) (2)
wherec is the speed of light RL is the fraction of the cavitythat is occupied by absorbing speciesαn(λj ) is the wave-length dependent absorption coefficient of the nth molecu-lar absorber andαcon(λj ) is the absorption coefficient dueto all other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16 September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Eq2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken fromVandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported bySaiz-Lopez et al(2004) (see alsoBall et al (2010)) The toppanel of Fig8 shows an example BBCRDS spectrum ob-tained during the campaign where the central and lower pan-els show respectively the I2 and NO2 contributions to themeasured absorption overlaid by their fitted reference spectrafrom the DOAS fitting routine During the RHaMBLe cam-paign the precision of the spectral retrievals was typically10 pptv (parts per trillion by volume) for I2 and 02 ppbv forNO2 (1σ uncertainty 304 s averaging time) Although notthe principal target of this deployment co-retrieval of theNO2 concentrations served as an important quality assuranceparameter with which to monitor the BBCRDS instrumentrsquosperformance Throughout the campaign the NO2 concen-trations measured by BBCRDS were in excellent quantita-tive agreement with NO2 measurements made by the Uni-versity of Yorkrsquos NOxy chemiluminescence instrument as
R J Leigh Bridging spatial scales in measurements of I2 7
α(λj) =RL
c
(1
τ(λj)minus 1τ0(λj)
)=sum
n
αn(λj) + αcon(λj) (2)
where c is the speed of light RL is the fraction of the cav-ity that is occupied by absorbing species αn(λj) is the wave-length dependent absorption coefficient of the nth molecularabsorber and αcon(λj) is the absorption coefficient due toall other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16th September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Equa-tion 2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2
were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken from Vandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported by Saiz-Lopez et al (2004) (see also Ball et al (2010)) The toppanel of Figure 8 shows an example BBCRDS spectrumobtained during the campaign where the central and lowerpanels show respectively the I2 and NO2 contributions tothe measured absorption overlaid by their fitted referencespectra from the DOAS fitting routine During the RHaM-BLe campaign the precision of the spectral retrievals wastypically 10 pptv for I2 and 02 ppbv for NO2 (1σ uncer-tainty 304 s averaging time) Although not the principaltarget of this deployment co-retrieval of the NO2 concentra-tions served as an important quality assurance parameter withwhich to monitor the BBCRDS instrumentrsquos performanceThroughout the campaign the NO2 concentrations measuredby BBCRDS were in excellent quantitative agreement withNO2 measurements made by the University of Yorkrsquos NOxy
chemiluminescence instrument as described in McFigganset al (2010) The NO2 amounts are also a valuable indicator
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14th September 2006 The upper panel shows the measuredspectrum (blue) after subtraction of the absorptions due to watervapour O4 and a second order polynomial accounting for the un-structured absorption contributions The red line shows a DOAS fitto the spectrumrsquos differential structure and the residual spectrumis shown in green The measured (blue) and fitted (red) absorptioncontributions due to I2 and NO2 are shown in the middle and lowerpanels respectively
of the possible extent of I2 recycling via IONO2 chemistryin the semi-polluted environment around Roscoff and so theNO2 field observations from both instruments are shown to-gether in the figures illustrating the measured and modelledI2 concentrations (see Figs 10 12 16 18) The generallygood agreement between the BBCRDS and chemilumines-cence measurements across a wide range of rapidly varyingNO2 concentrations is exemplified by the data from 14th-15th September shown in the bottom panel of Figure 10the gradient of a correlation plot of the NO2 concentrationsrecorded by the two instruments was 098 plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathDifferential Optical Absorption Spectroscopy (LP-DOAS)technique (Plane and Saiz-Lopez 2006) was used to mea-
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14 September 2006 The upper panel shows the measured spec-trum (blue) after subtraction of the absorptions due to water vapourO4 and a second order polynomial accounting for the unstructuredabsorption contributions The red line shows a DOAS fit to the spec-trumrsquos differential structure and the residual spectrum is shown ingreen The measured (blue) and fitted (red) absorption contribu-tions due to I2 and NO2 are shown in the middle and lower panelsrespectively
described inMcFiggans et al(2010) The NO2 amountsare also a valuable indicator of the possible extent of I2recycling via IONO2 chemistry in the semi-polluted envi-ronment around Roscoff and so the NO2 field observationsfrom both instruments are shown together in the figures il-lustrating the measured and modelled I2 concentrations (seeFigs10 12 16 and18) The generally good agreement be-tween the BBCRDS and chemiluminescence measurementsacross a wide range of rapidly varying NO2 concentrations isexemplified by the data from 14ndash15 September shown in thebottom panel of Fig 10 the gradient of a correlation plot ofthe NO2 concentrations recorded by the two instruments was098plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathdifferential optical absorption spectroscopy (LP-DOAS)
11830 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 9 Complete timeseries of model output from 5 to 28 Septem-ber 2006 From top down Regional emissions as calculated bythe model I2 concentrations at the measurement site based on foot-print analysis calculated with photolytic destruction and no recy-cling processes likewise but with recycling at 95 The lowesttwo panels show calculated I2 concentrations along the LP-DOASline of sight with photolytic destruction and no recycling processesand with recycling at 95
technique (Plane and Saiz-Lopez 2006) was used to mea-sure the concentrations of I2 OIO IO and NO3 The absorp-tion path extended 335 km from the SBR (48728 latitudeminus3988 longitude) to a small outcrop on the south west shoreof the Ile de Batz (4874 latitudeminus4036 longitude) wherea retroreflector array was placed to fold the optical path ndash seealso Fig1 The total optical path length was thus 67 km withthe beam 7 to 12 m above the mean sea level Full detailsof the DOAS instrument can be found elsewhere (Mahajanet al 2009 Saiz-Lopez and Plane 2004)
Briefly spectra were recorded with 025 nm resolution be-fore being converted into differential optical density spec-tra The contributions of individual absorbing species to the
Fig 10 Modelled and measured data from 13 to 15 September2006 The top panel shows BBCRDS data (red points and errorbars) with modelled concentrations of I2 at the site assuming 95recycling of I2 photolysed during the daytime (orange line) Themiddle panel shows LP-DOAS data (dark blue points and error bars)with modelled concentrations of I2 in the LP-DOAS light path as-suming 95 I2 recycling (blue line) The grey areas in the uppertwo plots indicate the range of modelled I2 values using recyclingassumptions fromR = 90 to 98 The photolysis frequency of I2is indicated by the green line in the upper two plots and tide by theblack line The bottom panel shows NO2 measured by the NOxychemiluminescence (black) and BBCRDS instruments (red pointsand error bars)
measured spectrum were determined by simultaneous fittingof their molecular absorption cross sections using singularvalue decomposition (Plane and Saiz-Lopez 2006) Aver-aged I2 concentrations along the line of sight were retrievedin the 535minus575 nm window on a number of days and nightsusing the I2 absorption cross sections ofSaiz-Lopez et al(2004) The full data set from the LP-DOAS instrumentis presented inMahajan et al(2009) andMcFiggans et al(2010)
For the present work footprints for the LP-DOAS instru-ment were calculated using the same footprint model (as-suming an 8 m height for the LP-DOAS light beam) withmodelled I2 amounts averaged for the footprints along theline of sight In this way the model provides a path lengthaveraged measurement of I2 along the LP-DOAS light pathwhich sampled emissions from a significant proportion of the
R J Leigh et al Bridging spatial scales in measurements of I2 11831
channel between the SBR and the Ile de Batz (see for exam-ple the footprints in Figs5 and6)
5 Calculation of total emissions and modelled I2 at thesite and along the LP DOAS line of sight
Spatially resolved emissions were calculated for each gridsquare within the model for each one minute time step Siteand LP-DOAS footprints were calculated taking account ofwind speed wind direction and tidal height The time takenfor I2 emissions to travel from their source within the foot-print to the detector was estimated from the windspeed mea-sured at the site In addition to dilution of the emitted I2(accounted for in the footprint) the model also consideredthe photolytic destruction of I2 during its transit to the mea-surement location using equation3
[I2](t)=[I2](0)ej I2middott (3)
where [I2](t) is the volume mixing ratio of I2 at time t and[I2](0) is the volume mixing ratio of I2 at time 0 directlyabove the emission source jI2 is the photolysis frequencyof I2 as measured by a spectral radiometer
Although no chemical modelling was attempted along thelines ofMahajan et al(2009) a simple recycling parameterR was also included in this work to mimic the effects of I2recycling via IONO2 chemistry downstream of the emissionsource This recycling was achieved through modification ofthe I2 photolytic destruction process to
[I2](t)=[I2](0)ej I2middot(1minusR)middott (4)
The recycling parameterR effectively permits a propor-tion of photodissociated I2 to be instantly reformed in ourmodel In order to test this approach the chemical model pre-sented previously inMahajan et al(2009) was re-run in thisstudy to examine the decay of I2 concentrations downwindof an emission source The chemical model was run twiceonce in the presence of NOx (for an assumed baseline NO2concentration of 2 ppbv) and again in the absence of NOxbut for reduced I2 photolysis rates (whilst keeping the pho-tolysis rates of other photolabile species unchanged) Goodqualitative and reasonable quantitative agreement was foundbetween these two scenarios when the I2 photolysis rate wasdecreased to 10 of its typical daytime value providing avalidation for our simplistic approach of reducing the effec-tive I2 photolysis rates to mimic the effect of I2 recycling viaIONO2 formation in this semi-polluted atmosphere For thepresent modelling study the recycling parameter was set to adefault value ofR = 095 ie 95 of the I2 that is photolysedis reformed by subsequent chemistry This is higher than theR = 090 suggested by theMahajan et al(2009) chemicalmodelling but tended to produce the best agreement withthe observational data For comparison modelled I2 concen-trations were also calculated for smaller (R = 090) and larger
(R=098) recycling efficiencies appropriate for respectivelyNO2 concentrations at and above the NO2 = 2 ppbv baselinecase of theMahajan et al(2009) model These results areadditionally shown in Fig 10 and the following other resultsfigures
Figure 9 shows total calculated regional emissions themodelled I2 mixing ratios in air advected to the measurementsite (BBCRDS) and the mean I2 mixing ratio in the air sam-pled along the LP-DOAS line of sight Modelled emissionsin the middle and bottom panels assume 95 recycling ofphotolysed I2 during the daytime Fig9 considers two I2 lossscenarios (i) dilution and irreversible photolytic loss accord-ing to Eq (3) and (ii) dilution and photolytic loss less a re-cycling assumption atR = 095 as per Eq (4) The modelledand measured I2 concentrations are compared in the follow-ing section
6 Comparison of modelled and measured I2
Owing to the challenges inherent in operating the BBCRDSand LP-DOAS instruments in the field and the requirementsfor measurements of other species to be taken by the sameinstrumentation I2 observations are unfortunately not avail-able throughout the campaign Results from three intensivemeasurement periods from each instrument are detailed be-low
61 Comparisons during the night
During the night in the absence of photolytic destructionI2 can be considered as a passive tracer to establish the linkbetween emissions and measurements Agreement betweenthe model and BBCRDS and LP-DOAS measurements atnight indicates acceptable model parameterisation of emis-sion rates seaweed spatial distribution meteorology and di-lution and dispersion within the emission footprint This cantherefore be used as a baseline for investigations during theday when additionally photolytic destruction and chemicalrecycling of I2 become important
Figure10 illustrates night-time measurements and modeldata from 13 to 15 September while Fig11 shows the cor-responding sources of I2 (differentiated by seaweed species)modelled at the measurement site and along the LP-DOASlight path In Fig10 the diurnal cycle is indicated by thephotolysis frequency of I2 (green line) with tide heightsshown in black During the period described in Fig10 therewere night-time low tides shortly after midnight on 14 and15 September 2006 when respectively both the LP-DOASand BBCRDS observed I2 substantially above their detec-tion limits and the model predicted peak I2 concentrations of50ndash100 pptv
Three modelled I2 datasets are presented in Fig10 (andother following data figures) covering the range of I2 re-cycling assumptions discussed in the previous section I2
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
R J Leigh et al Bridging spatial scales in measurements of I2 11825
2 R J Leigh Bridging spatial scales in measurements of I2
2 The Model
The present model incorporates two horizontal spatial scalesand a temporal domain with an additional vertical com-ponent included in footprint modelling calculations Thehorizontal grid consists of 746 by 227 elements each of00005times00005 degrees extending from -42075 to -3835degrees longitude and 486725 to 487855 degrees latitudeIn the Roscoff region this resolution corresponds to gridboxes of approximately 367 m longitudinally by 556 m lat-itudinally Bathymetry and macroalgal distribution informa-tion was mapped on to this model grid Tide and meteo-rological data was applied to this spatial information at 1-minute resolution from 5th to 28th September 2006 duringthe RHaMBLe campaign
21 Seaweed speciation and site bathymetry
The Roscoff inter-tidal zone in front of the SBR extendsmore than five kilometers in length and about 1 kilometerin width Mapping of the seaweed beds in the vicinity ofRoscoff has been attempted in two main studies One inthe early 1970s (Braud 1974) combined aerial photographsand in situ observations obtained from diving and field mea-surements The second study in the 1990s used both fieldand airborne spectrometers to map the seaweed and sea-grass beds near Roscoff (Bajjouk et al 1996) The pub-lished maps from these previous studies were used to con-struct a dedicated map for this work which was further vali-dated by field observations from September 2006 to Septem-ber 2009 This seaweed map was then superposed with abathymetry map of the area provided by L Leveque from theService Mer et Observation (Roscoff) The resulting mappeddistributions of Laminaria digitata Laminaria hyperboreaLaminaria ochroleuca Saccharina latisima Fucus andAscophyllum are shown in Figure 1 Ascophyllum and Fucusbeds are inherently mixed and are mapped together a con-stant mixing ratio of 6535 for AscophyllumFucus was usedin modelling their emissions
The vertical zonation of seaweed species is very distincton rocky shores with each species often forming a belt at acertain elevation in the eulittoral zone (the area between thehighest and the lowest tides) and also in the subtidal zone(the area extending below the zero of the marine charts) Itis thought that the driving force of this zonation is a com-bination of biotic factors and the tolerance of the differentspecies to abiotic factors such as temperature light salinitydehydration and mechanical forces caused by wave action(Luning 1990) A typical kelp bed from the Roscoff regionis shown in Figure 2
In the North Atlantic as exemplified in the study site infront of the SBR the eulittoral zone in sheltered habitatsis dominated both in coverage and biomass by brown algalspecies of the order of Fucales (fucoids) such as Fucus sppand Ascophyllum nodosum In addition four species of the
Fig 1 Bathymetry map and algal distributions used as inputs forthis modelling work The following key locations are marked themain measurement site (A) the LP-DOAS telescope (B) and theLP-DOAS retroreflector (C) Seaweed species are coded as followsL hyperborea - purple L digitata - green L ochroleuca - orangeSaccharina latissima - yellow AscophyllumFucus - red
Fig 2 The kelp bed at the Rocher du Loup at a low tide ofabout 05 m dominated by L digitata (kelp) and just above a beltof Himanthalia elongata (fucoid)
Fig 2 The kelp bed at the Rocher du Loup at a low tide ofabout 05 m dominated byL digitata (kelp) and just above a beltof Himanthalia elongata(fucoid)
The shallow inter-tidal zone at Roscoff results in the wa-ters closest to the shoreline being too shallow for Laminari-ales species So although a large horizontal surface area ofseaweed beds becomes exposed at low tide immediately infront of the SBR this mainly consists of fucoids (see Fig1)The distribution of seaweed species is rather patchy in theinter-tidal zone and is mainly dominated byFucus sppandAscophyllumbeds however there is also a small amount ofLdigitataSaccharina latissimaandL ochroleucain the chan-nel and tide pools between the site and the Ile de Batz islandto the north The south shore of Ile de Batz includes shel-tered shallow patchy habitats with sand and gravel which sur-round rocky areas covered by fucoids except where exposedto strong tide currentsLaminaria(mainly hyperborea) bedsextend to the north of Ile de Batz whereasL digitata flour-ishes in moderately exposed areas or at sites with strong wa-ter currents in the western part of the study site (Ile de Batzand islets west of Perharidy) and north-east from the Ile deBatzL digitataalso occurs in rockpools up to mid-tide leveland higher on wave-exposed coasts of the Ile de Batz
The average biomass densities in Table1 were obtainedfrom recent studies onAscophyllum nodosumat Roscoff(Gollety et al 2008) andL digitata (Gevaert et al 2008)and from the extensive long-term survey ofLaminarialespopulations by Ifremer (Arzel 1998) using the averagebiomass ofL digitata in September over the last ten yearsThe average volumic mass of each species was determinedexperimentally by filling a one-litre volume with seaweedthalli and determining the fresh weight of 5 replicates Thedepth limits of the various species at Roscoff were obtainedfrom previous mapping studies and from the Service Meret Observation in agreement with published data (Luning1990)
Fig 3 Emission rates assumed for the six species of macroalgae asa function of time following exposure to the atmosphere
22 Emission rates from exposed macroalgae
I2 emission rates for each species of macroalgae were esti-mated from the time since each model grid square was firstexposed to air by the changing tide and from the propen-sity of each seaweed species to emit For the former eachseaweed species was given a height attribute in the modelin order to account for the variable length and structure ofthe plants and therefore the variable water height at whichthe seaweed first breaks the water surface and becomes ex-posed to air These heights are shown in Table1 and re-sult in the I2 modelled emissions starting slightly in ad-vance of the surrounding sea bed itself becoming exposedSpecies specific I2 emission rates (in picomoles per minuteper gramme fresh weight) were parameterised from the lab-oratory study ofBall et al (2010) Figure3 shows the timedependent emission rates for each seaweed species used inthis work Fucusspecies andAscophyllumwere assumedto emit at a constant rate when exposed to air Emissionrates of theLaminaria species were assumed to decline af-ter first exposure to air with a common half life of 10 min-utes Laboratory studies have also found thatL digitatacansometimes resume bursts of strong I2 emission after havingbeen exposed to air for long periods (Dixneuf et al 2009Ball et al 2010) for simplicity here emission rates foreach species are assumed to become constant after 40 min-utes of continuous exposure Emissions from all speciescease immediately once the seaweed is re-covered by theincoming tide L ochroleuca for which measured emis-sion data were not available was assumed to emit at a rate
11826 R J Leigh et al Bridging spatial scales in measurements of I2
intermediate betweenL digitata and Saccharina latissimaThese emission rates were converted into emissions per m2
sea surface area using the assumptions shown in Table1 ofmass per m2 by species Emissions were assumed to mixinto an atmosphere layer of 15 cm depth providing a con-version into volume mixing ratio (VMR) This assumptionproduces peak VMRs immediately above the most strongly-emitting speciesL hyperboreaandL digitata of approxi-mately 15 ppbv (parts per billion by volume) immediately af-ter their first exposure to air consistent with the peak VMRsobserved byBall et al(2010) in their laboratory study
Actinic fluxes of solar radiation were measured using aMetcon spectral radiometer (Edwards and Monks 2003) andwere used to calculate the photolysis frequencies of a numberof trace gases including molecular iodine (jI2) ndash see Fig4The model temporal resolution was matched to the meteoro-logical dataset sampling of 1 min with all times within themodel expressed in Universal Time The model was appliedto data from 5 to 28 September 2006
23 Footprint analysis
Concentration footprints (as opposed to the more often usedflux footprints) were calculated for a range of wind veloc-ities and representative meteorological conditions using theanalytical approximation ofSchmid(1994) The model isa numerical solution to an analytical approximation of theadvection-diffusion equations The heterogeneity of the up-wind surface makes it rather difficult to draw firm conclu-sions about the exact form of the concentration footprintnevertheless the model is capable of providing sufficientlydetailed estimates for the purposes of this study
A surface roughness lengthz0 of 003 m was used wherethe fetch was across the inter-tidal zone For the water sur-faces encountered at high tidez0 was determined using therelationship described byZilitinkevich (1969)
z0= c1v
ulowast+
u2lowast
c2g(1)
whereulowast is the friction velocityg the acceleration due togravity v is the kinematic viscosity andc1 andc2 are coef-ficients with highly variable values estimated to be between00ndash048 (forc1) and frominfin to 811 (c2) In this studyintermediate values ofc1 = 01 andc2 = 320 were used Il-lustrations of footprint dimensions using this technique canbe found in Fig 13 ofMcFiggans et al(2010)
Using this technique footprints were calculated at fiveminute intervals throughout the campaign taking windspeedtide height and time of day as input parameters These inputparameters are shown in Fig4 Emission footprints werealso calculated atplusmn5 degrees either side of the wind direction(as measured at the SBR site) and the total modelled foot-print was taken to be the mean over all three of these footprintcalculations This averaging procedure aims to compensatefor temporal variability in wind direction within each 5 min
Fig 4 Tide I2 photolysis frequency and meteorological data usedto drive the model
time bin of the footprint calculation and for using wind datameasured at the SBR site to infer wind directionspeed overthe seaweed beds
Footprints calculated for each different tide height andwind strength were used in the model to characterise thetransport of I2 emissions to the site and into the LP DOAS in-strumentrsquos line of sight Following rotation appropriate to thewind direction the footprint for each time step was applied tothe emissions grid to estimate its contribution to the I2 con-centration observed by the BBCRDS and LP-DOAS instru-ments Footprints for the LP-DOAS were obtained throughintegration of all ldquopointrdquo footprints along the line of sight
3 Model results
The exposed regions of macroalgae were calculated for eachmodel time step and the resultant I2 emissions estimatedbased on the time since each grid cell had been exposed bythe tide and the seaweed species resident within the grid cellTwo example snap-shots of model footprints and emissionfields are shown in Figs5 and6 The curtain effect duringan ebb tide is shown in Fig5 as the initial exposure of theseaweed beds to air causes bursts of high I2 emissions espe-cially from the most potent emitters (red and green pixels)Emissions are lower and more uniform during the flow tide(see the dark blue pixels in Fig6)
Figure 7 shows the total regional emissions (panel 2)and the individual contributions from each seaweed species(panels 3ndash7) The greatest emissions correlated with the
R J Leigh et al Bridging spatial scales in measurements of I2 11827
Table 1 Derived bathymetry bands for each seaweed species included in this study with assumptions used to derive average fresh weightmass per m2 of coverage The final column details the height assumption for each species used to determine the tidal level at which seaweedbecomes exposed to the atmosphere
Species Minimum Maximum Average Average Average Averagedepth depth volumic mass volume biomass Height(m) (m) (kgFWm3) (m3m2) (kgFWm2) (m)
Fig 5 Model timestep from 1032 pm on 7 September 2006 dur-ing an ebb tide and one of the highest I2 concentrations predictedat site The wind speed and direction at this time were 565 msand 797 degrees respectively with the tide 085 m below the datumLP-DOAS and site footprints are shown in blue and grey shading re-spectively The modelled I2 emission fields are shown as red greenand dark blue pixels denoting emission rates of 1times1017 5times1016
and 25times1016molecules per grid square per second respectively
lowest low tides around 9ndash11 September and 23ndash25 Septem-ber when the largest area ofL digitata and particularlyL hyperboreabeds were uncovered The dominance of thesetwo Laminariaspecies in regional emissions is illustrated bythe the propensity of green and blue shading in panel 2 ofFig 7 these two species provide approximately 85 of to-tal regional emissions over the RHaMBLe campaign periodThe shapes of the emission profiles also change with the dif-ferent tidal spans The profiles are typically shorter and wideron days with the smallest tidal ranges around the middle ofthe campaign as seaweeds growing in shallow waters egfucusandascophyllumremain uncovered and thus contribut-ing their low level emissions throughout the majority of thetidal cycle In all cases the emission profiles are asymmetric
Fig 6 Model timestep from 548 pm on 14 September 2006 dur-ing a flow tide when the BBCRDS instrument measured signifi-cant concentrations of I2 The wind speed and direction at this timewere 19 ms and 309 degrees respectively with the tide 183 metresabove the datum The modelled I2 emission fields are shown on thesame scale as Fig5with dark blue purple and black pixels denotingemission rates of 25times1016 125times1016 andle 50times1015moleculesper grid square per second respectively The reduction in emissionssince first exposure can be seen with respect to Fig5
being biased towards greater emissions when the seaweedsare first uncovered by the retreating tide the initial burst ofemissions following first exposure is evident in the contribu-tions ofL ochroleucaand particularlySaccharina latissima(which grows in habitats spanning a narrow depth range)
4 I2 measurements during RHaMBLe
41 BBCRDS measurements
A broadband cavity ringdown spectrometer was deployedfrom a shipping container sited on the jetty in front of theSBR adjacent to the containers housing the campaignrsquos other
11828 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 7 Tide height (panel 1) and calculated regional emissions of I2 from 5 to 28 September 2006 (panel 2) The emission contributions aredifferentiated according to seaweed species by colour in panel 2 and are plotted separately in panels 3ndash7
in situ instruments (McFiggans et al 2010) Broadbandcavity ringdown spectroscopy (BBCRDS) uses light from apulsed broadband laser to measure the absorption spectrumof samples contained within a high finesse optical cavity(Bitter et al 2005 Ball and Jones 2003 2009) In this casethe BBCRDS instrument was configured to detect molecu-lar iodine using several of the I2 moleculersquos Blarr X absorp-tion bands in the wavelength range 560ndash570 nm Other at-mospheric gases (H2O NO2 and the oxygen dimer O4) alsoabsorb at these wavelengths and thus contribute to the mea-sured BBCRDS spectra
The BBCRDS system used for this study is based on aninstrument previously used to measure I2 at Mace Head (Ire-land) during the 2002 NAMBLEX campaign (Saiz-Lopezet al 2006 Heard et al 2006) as described in detail byBitteret al (2005) In the intervening years the instrumentrsquos per-formance has been enhanced significantly by upgrading sev-eral key components notably a new laser system that yieldspulsed broadband light with a factor of two wider bandwidth
at green wavelengths a new clocked CCD camera and im-proved analysis softwarespectral fitting routines A broad-band dye laser pumped by a 532 nm NdYAG laser (Sirah Co-bra and Surelight I-20 20 Hz repetition rate) generated lightpulses with an approximately Gaussian emission spectrumcentred at 563 nm (FWHM = 52 nm) This light was directedinto a 187 cm long ringdown cavity formed by two highlyreflective mirrors (Los Gatos peak reflectivity = 99993 at570 nm) Light exiting the ringdown cavity was collected andconveyed through a 100 microm core diameter fibre optic cable toan imaging spectrograph (Chromex 250is) where it was dis-persed in wavelength and imaged onto a clocked CCD cam-era (XCam CCDRem2) The time evolution of individualringdown events was recorded simultaneously at 512 differ-ent wavelengths one for each pixel row of the detector andlight from 50 ringdown events was integrated on the CCDcamera before storing the data to a computer Wavelengthresolved ringdown times were produced by fitting the ring-down decay in each pixel row (j = 1 to 512) The samplersquos
R J Leigh et al Bridging spatial scales in measurements of I2 11829
absorption spectrum was then calculated from sets of ring-down times measured when the cavity contained the sampleτ(λj ) and when flushed with dry nitrogenτ0(λj )
α(λj ) =RL
c
(1
τ(λj )minus
1
τ0(λj )
)=
sumn
αn(λj )+αcon(λj ) (2)
wherec is the speed of light RL is the fraction of the cavitythat is occupied by absorbing speciesαn(λj ) is the wave-length dependent absorption coefficient of the nth molecu-lar absorber andαcon(λj ) is the absorption coefficient dueto all other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16 September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Eq2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken fromVandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported bySaiz-Lopez et al(2004) (see alsoBall et al (2010)) The toppanel of Fig8 shows an example BBCRDS spectrum ob-tained during the campaign where the central and lower pan-els show respectively the I2 and NO2 contributions to themeasured absorption overlaid by their fitted reference spectrafrom the DOAS fitting routine During the RHaMBLe cam-paign the precision of the spectral retrievals was typically10 pptv (parts per trillion by volume) for I2 and 02 ppbv forNO2 (1σ uncertainty 304 s averaging time) Although notthe principal target of this deployment co-retrieval of theNO2 concentrations served as an important quality assuranceparameter with which to monitor the BBCRDS instrumentrsquosperformance Throughout the campaign the NO2 concen-trations measured by BBCRDS were in excellent quantita-tive agreement with NO2 measurements made by the Uni-versity of Yorkrsquos NOxy chemiluminescence instrument as
R J Leigh Bridging spatial scales in measurements of I2 7
α(λj) =RL
c
(1
τ(λj)minus 1τ0(λj)
)=sum
n
αn(λj) + αcon(λj) (2)
where c is the speed of light RL is the fraction of the cav-ity that is occupied by absorbing species αn(λj) is the wave-length dependent absorption coefficient of the nth molecularabsorber and αcon(λj) is the absorption coefficient due toall other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16th September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Equa-tion 2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2
were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken from Vandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported by Saiz-Lopez et al (2004) (see also Ball et al (2010)) The toppanel of Figure 8 shows an example BBCRDS spectrumobtained during the campaign where the central and lowerpanels show respectively the I2 and NO2 contributions tothe measured absorption overlaid by their fitted referencespectra from the DOAS fitting routine During the RHaM-BLe campaign the precision of the spectral retrievals wastypically 10 pptv for I2 and 02 ppbv for NO2 (1σ uncer-tainty 304 s averaging time) Although not the principaltarget of this deployment co-retrieval of the NO2 concentra-tions served as an important quality assurance parameter withwhich to monitor the BBCRDS instrumentrsquos performanceThroughout the campaign the NO2 concentrations measuredby BBCRDS were in excellent quantitative agreement withNO2 measurements made by the University of Yorkrsquos NOxy
chemiluminescence instrument as described in McFigganset al (2010) The NO2 amounts are also a valuable indicator
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14th September 2006 The upper panel shows the measuredspectrum (blue) after subtraction of the absorptions due to watervapour O4 and a second order polynomial accounting for the un-structured absorption contributions The red line shows a DOAS fitto the spectrumrsquos differential structure and the residual spectrumis shown in green The measured (blue) and fitted (red) absorptioncontributions due to I2 and NO2 are shown in the middle and lowerpanels respectively
of the possible extent of I2 recycling via IONO2 chemistryin the semi-polluted environment around Roscoff and so theNO2 field observations from both instruments are shown to-gether in the figures illustrating the measured and modelledI2 concentrations (see Figs 10 12 16 18) The generallygood agreement between the BBCRDS and chemilumines-cence measurements across a wide range of rapidly varyingNO2 concentrations is exemplified by the data from 14th-15th September shown in the bottom panel of Figure 10the gradient of a correlation plot of the NO2 concentrationsrecorded by the two instruments was 098 plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathDifferential Optical Absorption Spectroscopy (LP-DOAS)technique (Plane and Saiz-Lopez 2006) was used to mea-
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14 September 2006 The upper panel shows the measured spec-trum (blue) after subtraction of the absorptions due to water vapourO4 and a second order polynomial accounting for the unstructuredabsorption contributions The red line shows a DOAS fit to the spec-trumrsquos differential structure and the residual spectrum is shown ingreen The measured (blue) and fitted (red) absorption contribu-tions due to I2 and NO2 are shown in the middle and lower panelsrespectively
described inMcFiggans et al(2010) The NO2 amountsare also a valuable indicator of the possible extent of I2recycling via IONO2 chemistry in the semi-polluted envi-ronment around Roscoff and so the NO2 field observationsfrom both instruments are shown together in the figures il-lustrating the measured and modelled I2 concentrations (seeFigs10 12 16 and18) The generally good agreement be-tween the BBCRDS and chemiluminescence measurementsacross a wide range of rapidly varying NO2 concentrations isexemplified by the data from 14ndash15 September shown in thebottom panel of Fig 10 the gradient of a correlation plot ofthe NO2 concentrations recorded by the two instruments was098plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathdifferential optical absorption spectroscopy (LP-DOAS)
11830 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 9 Complete timeseries of model output from 5 to 28 Septem-ber 2006 From top down Regional emissions as calculated bythe model I2 concentrations at the measurement site based on foot-print analysis calculated with photolytic destruction and no recy-cling processes likewise but with recycling at 95 The lowesttwo panels show calculated I2 concentrations along the LP-DOASline of sight with photolytic destruction and no recycling processesand with recycling at 95
technique (Plane and Saiz-Lopez 2006) was used to mea-sure the concentrations of I2 OIO IO and NO3 The absorp-tion path extended 335 km from the SBR (48728 latitudeminus3988 longitude) to a small outcrop on the south west shoreof the Ile de Batz (4874 latitudeminus4036 longitude) wherea retroreflector array was placed to fold the optical path ndash seealso Fig1 The total optical path length was thus 67 km withthe beam 7 to 12 m above the mean sea level Full detailsof the DOAS instrument can be found elsewhere (Mahajanet al 2009 Saiz-Lopez and Plane 2004)
Briefly spectra were recorded with 025 nm resolution be-fore being converted into differential optical density spec-tra The contributions of individual absorbing species to the
Fig 10 Modelled and measured data from 13 to 15 September2006 The top panel shows BBCRDS data (red points and errorbars) with modelled concentrations of I2 at the site assuming 95recycling of I2 photolysed during the daytime (orange line) Themiddle panel shows LP-DOAS data (dark blue points and error bars)with modelled concentrations of I2 in the LP-DOAS light path as-suming 95 I2 recycling (blue line) The grey areas in the uppertwo plots indicate the range of modelled I2 values using recyclingassumptions fromR = 90 to 98 The photolysis frequency of I2is indicated by the green line in the upper two plots and tide by theblack line The bottom panel shows NO2 measured by the NOxychemiluminescence (black) and BBCRDS instruments (red pointsand error bars)
measured spectrum were determined by simultaneous fittingof their molecular absorption cross sections using singularvalue decomposition (Plane and Saiz-Lopez 2006) Aver-aged I2 concentrations along the line of sight were retrievedin the 535minus575 nm window on a number of days and nightsusing the I2 absorption cross sections ofSaiz-Lopez et al(2004) The full data set from the LP-DOAS instrumentis presented inMahajan et al(2009) andMcFiggans et al(2010)
For the present work footprints for the LP-DOAS instru-ment were calculated using the same footprint model (as-suming an 8 m height for the LP-DOAS light beam) withmodelled I2 amounts averaged for the footprints along theline of sight In this way the model provides a path lengthaveraged measurement of I2 along the LP-DOAS light pathwhich sampled emissions from a significant proportion of the
R J Leigh et al Bridging spatial scales in measurements of I2 11831
channel between the SBR and the Ile de Batz (see for exam-ple the footprints in Figs5 and6)
5 Calculation of total emissions and modelled I2 at thesite and along the LP DOAS line of sight
Spatially resolved emissions were calculated for each gridsquare within the model for each one minute time step Siteand LP-DOAS footprints were calculated taking account ofwind speed wind direction and tidal height The time takenfor I2 emissions to travel from their source within the foot-print to the detector was estimated from the windspeed mea-sured at the site In addition to dilution of the emitted I2(accounted for in the footprint) the model also consideredthe photolytic destruction of I2 during its transit to the mea-surement location using equation3
[I2](t)=[I2](0)ej I2middott (3)
where [I2](t) is the volume mixing ratio of I2 at time t and[I2](0) is the volume mixing ratio of I2 at time 0 directlyabove the emission source jI2 is the photolysis frequencyof I2 as measured by a spectral radiometer
Although no chemical modelling was attempted along thelines ofMahajan et al(2009) a simple recycling parameterR was also included in this work to mimic the effects of I2recycling via IONO2 chemistry downstream of the emissionsource This recycling was achieved through modification ofthe I2 photolytic destruction process to
[I2](t)=[I2](0)ej I2middot(1minusR)middott (4)
The recycling parameterR effectively permits a propor-tion of photodissociated I2 to be instantly reformed in ourmodel In order to test this approach the chemical model pre-sented previously inMahajan et al(2009) was re-run in thisstudy to examine the decay of I2 concentrations downwindof an emission source The chemical model was run twiceonce in the presence of NOx (for an assumed baseline NO2concentration of 2 ppbv) and again in the absence of NOxbut for reduced I2 photolysis rates (whilst keeping the pho-tolysis rates of other photolabile species unchanged) Goodqualitative and reasonable quantitative agreement was foundbetween these two scenarios when the I2 photolysis rate wasdecreased to 10 of its typical daytime value providing avalidation for our simplistic approach of reducing the effec-tive I2 photolysis rates to mimic the effect of I2 recycling viaIONO2 formation in this semi-polluted atmosphere For thepresent modelling study the recycling parameter was set to adefault value ofR = 095 ie 95 of the I2 that is photolysedis reformed by subsequent chemistry This is higher than theR = 090 suggested by theMahajan et al(2009) chemicalmodelling but tended to produce the best agreement withthe observational data For comparison modelled I2 concen-trations were also calculated for smaller (R = 090) and larger
(R=098) recycling efficiencies appropriate for respectivelyNO2 concentrations at and above the NO2 = 2 ppbv baselinecase of theMahajan et al(2009) model These results areadditionally shown in Fig 10 and the following other resultsfigures
Figure 9 shows total calculated regional emissions themodelled I2 mixing ratios in air advected to the measurementsite (BBCRDS) and the mean I2 mixing ratio in the air sam-pled along the LP-DOAS line of sight Modelled emissionsin the middle and bottom panels assume 95 recycling ofphotolysed I2 during the daytime Fig9 considers two I2 lossscenarios (i) dilution and irreversible photolytic loss accord-ing to Eq (3) and (ii) dilution and photolytic loss less a re-cycling assumption atR = 095 as per Eq (4) The modelledand measured I2 concentrations are compared in the follow-ing section
6 Comparison of modelled and measured I2
Owing to the challenges inherent in operating the BBCRDSand LP-DOAS instruments in the field and the requirementsfor measurements of other species to be taken by the sameinstrumentation I2 observations are unfortunately not avail-able throughout the campaign Results from three intensivemeasurement periods from each instrument are detailed be-low
61 Comparisons during the night
During the night in the absence of photolytic destructionI2 can be considered as a passive tracer to establish the linkbetween emissions and measurements Agreement betweenthe model and BBCRDS and LP-DOAS measurements atnight indicates acceptable model parameterisation of emis-sion rates seaweed spatial distribution meteorology and di-lution and dispersion within the emission footprint This cantherefore be used as a baseline for investigations during theday when additionally photolytic destruction and chemicalrecycling of I2 become important
Figure10 illustrates night-time measurements and modeldata from 13 to 15 September while Fig11 shows the cor-responding sources of I2 (differentiated by seaweed species)modelled at the measurement site and along the LP-DOASlight path In Fig10 the diurnal cycle is indicated by thephotolysis frequency of I2 (green line) with tide heightsshown in black During the period described in Fig10 therewere night-time low tides shortly after midnight on 14 and15 September 2006 when respectively both the LP-DOASand BBCRDS observed I2 substantially above their detec-tion limits and the model predicted peak I2 concentrations of50ndash100 pptv
Three modelled I2 datasets are presented in Fig10 (andother following data figures) covering the range of I2 re-cycling assumptions discussed in the previous section I2
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
11826 R J Leigh et al Bridging spatial scales in measurements of I2
intermediate betweenL digitata and Saccharina latissimaThese emission rates were converted into emissions per m2
sea surface area using the assumptions shown in Table1 ofmass per m2 by species Emissions were assumed to mixinto an atmosphere layer of 15 cm depth providing a con-version into volume mixing ratio (VMR) This assumptionproduces peak VMRs immediately above the most strongly-emitting speciesL hyperboreaandL digitata of approxi-mately 15 ppbv (parts per billion by volume) immediately af-ter their first exposure to air consistent with the peak VMRsobserved byBall et al(2010) in their laboratory study
Actinic fluxes of solar radiation were measured using aMetcon spectral radiometer (Edwards and Monks 2003) andwere used to calculate the photolysis frequencies of a numberof trace gases including molecular iodine (jI2) ndash see Fig4The model temporal resolution was matched to the meteoro-logical dataset sampling of 1 min with all times within themodel expressed in Universal Time The model was appliedto data from 5 to 28 September 2006
23 Footprint analysis
Concentration footprints (as opposed to the more often usedflux footprints) were calculated for a range of wind veloc-ities and representative meteorological conditions using theanalytical approximation ofSchmid(1994) The model isa numerical solution to an analytical approximation of theadvection-diffusion equations The heterogeneity of the up-wind surface makes it rather difficult to draw firm conclu-sions about the exact form of the concentration footprintnevertheless the model is capable of providing sufficientlydetailed estimates for the purposes of this study
A surface roughness lengthz0 of 003 m was used wherethe fetch was across the inter-tidal zone For the water sur-faces encountered at high tidez0 was determined using therelationship described byZilitinkevich (1969)
z0= c1v
ulowast+
u2lowast
c2g(1)
whereulowast is the friction velocityg the acceleration due togravity v is the kinematic viscosity andc1 andc2 are coef-ficients with highly variable values estimated to be between00ndash048 (forc1) and frominfin to 811 (c2) In this studyintermediate values ofc1 = 01 andc2 = 320 were used Il-lustrations of footprint dimensions using this technique canbe found in Fig 13 ofMcFiggans et al(2010)
Using this technique footprints were calculated at fiveminute intervals throughout the campaign taking windspeedtide height and time of day as input parameters These inputparameters are shown in Fig4 Emission footprints werealso calculated atplusmn5 degrees either side of the wind direction(as measured at the SBR site) and the total modelled foot-print was taken to be the mean over all three of these footprintcalculations This averaging procedure aims to compensatefor temporal variability in wind direction within each 5 min
Fig 4 Tide I2 photolysis frequency and meteorological data usedto drive the model
time bin of the footprint calculation and for using wind datameasured at the SBR site to infer wind directionspeed overthe seaweed beds
Footprints calculated for each different tide height andwind strength were used in the model to characterise thetransport of I2 emissions to the site and into the LP DOAS in-strumentrsquos line of sight Following rotation appropriate to thewind direction the footprint for each time step was applied tothe emissions grid to estimate its contribution to the I2 con-centration observed by the BBCRDS and LP-DOAS instru-ments Footprints for the LP-DOAS were obtained throughintegration of all ldquopointrdquo footprints along the line of sight
3 Model results
The exposed regions of macroalgae were calculated for eachmodel time step and the resultant I2 emissions estimatedbased on the time since each grid cell had been exposed bythe tide and the seaweed species resident within the grid cellTwo example snap-shots of model footprints and emissionfields are shown in Figs5 and6 The curtain effect duringan ebb tide is shown in Fig5 as the initial exposure of theseaweed beds to air causes bursts of high I2 emissions espe-cially from the most potent emitters (red and green pixels)Emissions are lower and more uniform during the flow tide(see the dark blue pixels in Fig6)
Figure 7 shows the total regional emissions (panel 2)and the individual contributions from each seaweed species(panels 3ndash7) The greatest emissions correlated with the
R J Leigh et al Bridging spatial scales in measurements of I2 11827
Table 1 Derived bathymetry bands for each seaweed species included in this study with assumptions used to derive average fresh weightmass per m2 of coverage The final column details the height assumption for each species used to determine the tidal level at which seaweedbecomes exposed to the atmosphere
Species Minimum Maximum Average Average Average Averagedepth depth volumic mass volume biomass Height(m) (m) (kgFWm3) (m3m2) (kgFWm2) (m)
Fig 5 Model timestep from 1032 pm on 7 September 2006 dur-ing an ebb tide and one of the highest I2 concentrations predictedat site The wind speed and direction at this time were 565 msand 797 degrees respectively with the tide 085 m below the datumLP-DOAS and site footprints are shown in blue and grey shading re-spectively The modelled I2 emission fields are shown as red greenand dark blue pixels denoting emission rates of 1times1017 5times1016
and 25times1016molecules per grid square per second respectively
lowest low tides around 9ndash11 September and 23ndash25 Septem-ber when the largest area ofL digitata and particularlyL hyperboreabeds were uncovered The dominance of thesetwo Laminariaspecies in regional emissions is illustrated bythe the propensity of green and blue shading in panel 2 ofFig 7 these two species provide approximately 85 of to-tal regional emissions over the RHaMBLe campaign periodThe shapes of the emission profiles also change with the dif-ferent tidal spans The profiles are typically shorter and wideron days with the smallest tidal ranges around the middle ofthe campaign as seaweeds growing in shallow waters egfucusandascophyllumremain uncovered and thus contribut-ing their low level emissions throughout the majority of thetidal cycle In all cases the emission profiles are asymmetric
Fig 6 Model timestep from 548 pm on 14 September 2006 dur-ing a flow tide when the BBCRDS instrument measured signifi-cant concentrations of I2 The wind speed and direction at this timewere 19 ms and 309 degrees respectively with the tide 183 metresabove the datum The modelled I2 emission fields are shown on thesame scale as Fig5with dark blue purple and black pixels denotingemission rates of 25times1016 125times1016 andle 50times1015moleculesper grid square per second respectively The reduction in emissionssince first exposure can be seen with respect to Fig5
being biased towards greater emissions when the seaweedsare first uncovered by the retreating tide the initial burst ofemissions following first exposure is evident in the contribu-tions ofL ochroleucaand particularlySaccharina latissima(which grows in habitats spanning a narrow depth range)
4 I2 measurements during RHaMBLe
41 BBCRDS measurements
A broadband cavity ringdown spectrometer was deployedfrom a shipping container sited on the jetty in front of theSBR adjacent to the containers housing the campaignrsquos other
11828 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 7 Tide height (panel 1) and calculated regional emissions of I2 from 5 to 28 September 2006 (panel 2) The emission contributions aredifferentiated according to seaweed species by colour in panel 2 and are plotted separately in panels 3ndash7
in situ instruments (McFiggans et al 2010) Broadbandcavity ringdown spectroscopy (BBCRDS) uses light from apulsed broadband laser to measure the absorption spectrumof samples contained within a high finesse optical cavity(Bitter et al 2005 Ball and Jones 2003 2009) In this casethe BBCRDS instrument was configured to detect molecu-lar iodine using several of the I2 moleculersquos Blarr X absorp-tion bands in the wavelength range 560ndash570 nm Other at-mospheric gases (H2O NO2 and the oxygen dimer O4) alsoabsorb at these wavelengths and thus contribute to the mea-sured BBCRDS spectra
The BBCRDS system used for this study is based on aninstrument previously used to measure I2 at Mace Head (Ire-land) during the 2002 NAMBLEX campaign (Saiz-Lopezet al 2006 Heard et al 2006) as described in detail byBitteret al (2005) In the intervening years the instrumentrsquos per-formance has been enhanced significantly by upgrading sev-eral key components notably a new laser system that yieldspulsed broadband light with a factor of two wider bandwidth
at green wavelengths a new clocked CCD camera and im-proved analysis softwarespectral fitting routines A broad-band dye laser pumped by a 532 nm NdYAG laser (Sirah Co-bra and Surelight I-20 20 Hz repetition rate) generated lightpulses with an approximately Gaussian emission spectrumcentred at 563 nm (FWHM = 52 nm) This light was directedinto a 187 cm long ringdown cavity formed by two highlyreflective mirrors (Los Gatos peak reflectivity = 99993 at570 nm) Light exiting the ringdown cavity was collected andconveyed through a 100 microm core diameter fibre optic cable toan imaging spectrograph (Chromex 250is) where it was dis-persed in wavelength and imaged onto a clocked CCD cam-era (XCam CCDRem2) The time evolution of individualringdown events was recorded simultaneously at 512 differ-ent wavelengths one for each pixel row of the detector andlight from 50 ringdown events was integrated on the CCDcamera before storing the data to a computer Wavelengthresolved ringdown times were produced by fitting the ring-down decay in each pixel row (j = 1 to 512) The samplersquos
R J Leigh et al Bridging spatial scales in measurements of I2 11829
absorption spectrum was then calculated from sets of ring-down times measured when the cavity contained the sampleτ(λj ) and when flushed with dry nitrogenτ0(λj )
α(λj ) =RL
c
(1
τ(λj )minus
1
τ0(λj )
)=
sumn
αn(λj )+αcon(λj ) (2)
wherec is the speed of light RL is the fraction of the cavitythat is occupied by absorbing speciesαn(λj ) is the wave-length dependent absorption coefficient of the nth molecu-lar absorber andαcon(λj ) is the absorption coefficient dueto all other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16 September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Eq2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken fromVandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported bySaiz-Lopez et al(2004) (see alsoBall et al (2010)) The toppanel of Fig8 shows an example BBCRDS spectrum ob-tained during the campaign where the central and lower pan-els show respectively the I2 and NO2 contributions to themeasured absorption overlaid by their fitted reference spectrafrom the DOAS fitting routine During the RHaMBLe cam-paign the precision of the spectral retrievals was typically10 pptv (parts per trillion by volume) for I2 and 02 ppbv forNO2 (1σ uncertainty 304 s averaging time) Although notthe principal target of this deployment co-retrieval of theNO2 concentrations served as an important quality assuranceparameter with which to monitor the BBCRDS instrumentrsquosperformance Throughout the campaign the NO2 concen-trations measured by BBCRDS were in excellent quantita-tive agreement with NO2 measurements made by the Uni-versity of Yorkrsquos NOxy chemiluminescence instrument as
R J Leigh Bridging spatial scales in measurements of I2 7
α(λj) =RL
c
(1
τ(λj)minus 1τ0(λj)
)=sum
n
αn(λj) + αcon(λj) (2)
where c is the speed of light RL is the fraction of the cav-ity that is occupied by absorbing species αn(λj) is the wave-length dependent absorption coefficient of the nth molecularabsorber and αcon(λj) is the absorption coefficient due toall other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16th September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Equa-tion 2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2
were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken from Vandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported by Saiz-Lopez et al (2004) (see also Ball et al (2010)) The toppanel of Figure 8 shows an example BBCRDS spectrumobtained during the campaign where the central and lowerpanels show respectively the I2 and NO2 contributions tothe measured absorption overlaid by their fitted referencespectra from the DOAS fitting routine During the RHaM-BLe campaign the precision of the spectral retrievals wastypically 10 pptv for I2 and 02 ppbv for NO2 (1σ uncer-tainty 304 s averaging time) Although not the principaltarget of this deployment co-retrieval of the NO2 concentra-tions served as an important quality assurance parameter withwhich to monitor the BBCRDS instrumentrsquos performanceThroughout the campaign the NO2 concentrations measuredby BBCRDS were in excellent quantitative agreement withNO2 measurements made by the University of Yorkrsquos NOxy
chemiluminescence instrument as described in McFigganset al (2010) The NO2 amounts are also a valuable indicator
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14th September 2006 The upper panel shows the measuredspectrum (blue) after subtraction of the absorptions due to watervapour O4 and a second order polynomial accounting for the un-structured absorption contributions The red line shows a DOAS fitto the spectrumrsquos differential structure and the residual spectrumis shown in green The measured (blue) and fitted (red) absorptioncontributions due to I2 and NO2 are shown in the middle and lowerpanels respectively
of the possible extent of I2 recycling via IONO2 chemistryin the semi-polluted environment around Roscoff and so theNO2 field observations from both instruments are shown to-gether in the figures illustrating the measured and modelledI2 concentrations (see Figs 10 12 16 18) The generallygood agreement between the BBCRDS and chemilumines-cence measurements across a wide range of rapidly varyingNO2 concentrations is exemplified by the data from 14th-15th September shown in the bottom panel of Figure 10the gradient of a correlation plot of the NO2 concentrationsrecorded by the two instruments was 098 plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathDifferential Optical Absorption Spectroscopy (LP-DOAS)technique (Plane and Saiz-Lopez 2006) was used to mea-
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14 September 2006 The upper panel shows the measured spec-trum (blue) after subtraction of the absorptions due to water vapourO4 and a second order polynomial accounting for the unstructuredabsorption contributions The red line shows a DOAS fit to the spec-trumrsquos differential structure and the residual spectrum is shown ingreen The measured (blue) and fitted (red) absorption contribu-tions due to I2 and NO2 are shown in the middle and lower panelsrespectively
described inMcFiggans et al(2010) The NO2 amountsare also a valuable indicator of the possible extent of I2recycling via IONO2 chemistry in the semi-polluted envi-ronment around Roscoff and so the NO2 field observationsfrom both instruments are shown together in the figures il-lustrating the measured and modelled I2 concentrations (seeFigs10 12 16 and18) The generally good agreement be-tween the BBCRDS and chemiluminescence measurementsacross a wide range of rapidly varying NO2 concentrations isexemplified by the data from 14ndash15 September shown in thebottom panel of Fig 10 the gradient of a correlation plot ofthe NO2 concentrations recorded by the two instruments was098plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathdifferential optical absorption spectroscopy (LP-DOAS)
11830 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 9 Complete timeseries of model output from 5 to 28 Septem-ber 2006 From top down Regional emissions as calculated bythe model I2 concentrations at the measurement site based on foot-print analysis calculated with photolytic destruction and no recy-cling processes likewise but with recycling at 95 The lowesttwo panels show calculated I2 concentrations along the LP-DOASline of sight with photolytic destruction and no recycling processesand with recycling at 95
technique (Plane and Saiz-Lopez 2006) was used to mea-sure the concentrations of I2 OIO IO and NO3 The absorp-tion path extended 335 km from the SBR (48728 latitudeminus3988 longitude) to a small outcrop on the south west shoreof the Ile de Batz (4874 latitudeminus4036 longitude) wherea retroreflector array was placed to fold the optical path ndash seealso Fig1 The total optical path length was thus 67 km withthe beam 7 to 12 m above the mean sea level Full detailsof the DOAS instrument can be found elsewhere (Mahajanet al 2009 Saiz-Lopez and Plane 2004)
Briefly spectra were recorded with 025 nm resolution be-fore being converted into differential optical density spec-tra The contributions of individual absorbing species to the
Fig 10 Modelled and measured data from 13 to 15 September2006 The top panel shows BBCRDS data (red points and errorbars) with modelled concentrations of I2 at the site assuming 95recycling of I2 photolysed during the daytime (orange line) Themiddle panel shows LP-DOAS data (dark blue points and error bars)with modelled concentrations of I2 in the LP-DOAS light path as-suming 95 I2 recycling (blue line) The grey areas in the uppertwo plots indicate the range of modelled I2 values using recyclingassumptions fromR = 90 to 98 The photolysis frequency of I2is indicated by the green line in the upper two plots and tide by theblack line The bottom panel shows NO2 measured by the NOxychemiluminescence (black) and BBCRDS instruments (red pointsand error bars)
measured spectrum were determined by simultaneous fittingof their molecular absorption cross sections using singularvalue decomposition (Plane and Saiz-Lopez 2006) Aver-aged I2 concentrations along the line of sight were retrievedin the 535minus575 nm window on a number of days and nightsusing the I2 absorption cross sections ofSaiz-Lopez et al(2004) The full data set from the LP-DOAS instrumentis presented inMahajan et al(2009) andMcFiggans et al(2010)
For the present work footprints for the LP-DOAS instru-ment were calculated using the same footprint model (as-suming an 8 m height for the LP-DOAS light beam) withmodelled I2 amounts averaged for the footprints along theline of sight In this way the model provides a path lengthaveraged measurement of I2 along the LP-DOAS light pathwhich sampled emissions from a significant proportion of the
R J Leigh et al Bridging spatial scales in measurements of I2 11831
channel between the SBR and the Ile de Batz (see for exam-ple the footprints in Figs5 and6)
5 Calculation of total emissions and modelled I2 at thesite and along the LP DOAS line of sight
Spatially resolved emissions were calculated for each gridsquare within the model for each one minute time step Siteand LP-DOAS footprints were calculated taking account ofwind speed wind direction and tidal height The time takenfor I2 emissions to travel from their source within the foot-print to the detector was estimated from the windspeed mea-sured at the site In addition to dilution of the emitted I2(accounted for in the footprint) the model also consideredthe photolytic destruction of I2 during its transit to the mea-surement location using equation3
[I2](t)=[I2](0)ej I2middott (3)
where [I2](t) is the volume mixing ratio of I2 at time t and[I2](0) is the volume mixing ratio of I2 at time 0 directlyabove the emission source jI2 is the photolysis frequencyof I2 as measured by a spectral radiometer
Although no chemical modelling was attempted along thelines ofMahajan et al(2009) a simple recycling parameterR was also included in this work to mimic the effects of I2recycling via IONO2 chemistry downstream of the emissionsource This recycling was achieved through modification ofthe I2 photolytic destruction process to
[I2](t)=[I2](0)ej I2middot(1minusR)middott (4)
The recycling parameterR effectively permits a propor-tion of photodissociated I2 to be instantly reformed in ourmodel In order to test this approach the chemical model pre-sented previously inMahajan et al(2009) was re-run in thisstudy to examine the decay of I2 concentrations downwindof an emission source The chemical model was run twiceonce in the presence of NOx (for an assumed baseline NO2concentration of 2 ppbv) and again in the absence of NOxbut for reduced I2 photolysis rates (whilst keeping the pho-tolysis rates of other photolabile species unchanged) Goodqualitative and reasonable quantitative agreement was foundbetween these two scenarios when the I2 photolysis rate wasdecreased to 10 of its typical daytime value providing avalidation for our simplistic approach of reducing the effec-tive I2 photolysis rates to mimic the effect of I2 recycling viaIONO2 formation in this semi-polluted atmosphere For thepresent modelling study the recycling parameter was set to adefault value ofR = 095 ie 95 of the I2 that is photolysedis reformed by subsequent chemistry This is higher than theR = 090 suggested by theMahajan et al(2009) chemicalmodelling but tended to produce the best agreement withthe observational data For comparison modelled I2 concen-trations were also calculated for smaller (R = 090) and larger
(R=098) recycling efficiencies appropriate for respectivelyNO2 concentrations at and above the NO2 = 2 ppbv baselinecase of theMahajan et al(2009) model These results areadditionally shown in Fig 10 and the following other resultsfigures
Figure 9 shows total calculated regional emissions themodelled I2 mixing ratios in air advected to the measurementsite (BBCRDS) and the mean I2 mixing ratio in the air sam-pled along the LP-DOAS line of sight Modelled emissionsin the middle and bottom panels assume 95 recycling ofphotolysed I2 during the daytime Fig9 considers two I2 lossscenarios (i) dilution and irreversible photolytic loss accord-ing to Eq (3) and (ii) dilution and photolytic loss less a re-cycling assumption atR = 095 as per Eq (4) The modelledand measured I2 concentrations are compared in the follow-ing section
6 Comparison of modelled and measured I2
Owing to the challenges inherent in operating the BBCRDSand LP-DOAS instruments in the field and the requirementsfor measurements of other species to be taken by the sameinstrumentation I2 observations are unfortunately not avail-able throughout the campaign Results from three intensivemeasurement periods from each instrument are detailed be-low
61 Comparisons during the night
During the night in the absence of photolytic destructionI2 can be considered as a passive tracer to establish the linkbetween emissions and measurements Agreement betweenthe model and BBCRDS and LP-DOAS measurements atnight indicates acceptable model parameterisation of emis-sion rates seaweed spatial distribution meteorology and di-lution and dispersion within the emission footprint This cantherefore be used as a baseline for investigations during theday when additionally photolytic destruction and chemicalrecycling of I2 become important
Figure10 illustrates night-time measurements and modeldata from 13 to 15 September while Fig11 shows the cor-responding sources of I2 (differentiated by seaweed species)modelled at the measurement site and along the LP-DOASlight path In Fig10 the diurnal cycle is indicated by thephotolysis frequency of I2 (green line) with tide heightsshown in black During the period described in Fig10 therewere night-time low tides shortly after midnight on 14 and15 September 2006 when respectively both the LP-DOASand BBCRDS observed I2 substantially above their detec-tion limits and the model predicted peak I2 concentrations of50ndash100 pptv
Three modelled I2 datasets are presented in Fig10 (andother following data figures) covering the range of I2 re-cycling assumptions discussed in the previous section I2
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
R J Leigh et al Bridging spatial scales in measurements of I2 11827
Table 1 Derived bathymetry bands for each seaweed species included in this study with assumptions used to derive average fresh weightmass per m2 of coverage The final column details the height assumption for each species used to determine the tidal level at which seaweedbecomes exposed to the atmosphere
Species Minimum Maximum Average Average Average Averagedepth depth volumic mass volume biomass Height(m) (m) (kgFWm3) (m3m2) (kgFWm2) (m)
Fig 5 Model timestep from 1032 pm on 7 September 2006 dur-ing an ebb tide and one of the highest I2 concentrations predictedat site The wind speed and direction at this time were 565 msand 797 degrees respectively with the tide 085 m below the datumLP-DOAS and site footprints are shown in blue and grey shading re-spectively The modelled I2 emission fields are shown as red greenand dark blue pixels denoting emission rates of 1times1017 5times1016
and 25times1016molecules per grid square per second respectively
lowest low tides around 9ndash11 September and 23ndash25 Septem-ber when the largest area ofL digitata and particularlyL hyperboreabeds were uncovered The dominance of thesetwo Laminariaspecies in regional emissions is illustrated bythe the propensity of green and blue shading in panel 2 ofFig 7 these two species provide approximately 85 of to-tal regional emissions over the RHaMBLe campaign periodThe shapes of the emission profiles also change with the dif-ferent tidal spans The profiles are typically shorter and wideron days with the smallest tidal ranges around the middle ofthe campaign as seaweeds growing in shallow waters egfucusandascophyllumremain uncovered and thus contribut-ing their low level emissions throughout the majority of thetidal cycle In all cases the emission profiles are asymmetric
Fig 6 Model timestep from 548 pm on 14 September 2006 dur-ing a flow tide when the BBCRDS instrument measured signifi-cant concentrations of I2 The wind speed and direction at this timewere 19 ms and 309 degrees respectively with the tide 183 metresabove the datum The modelled I2 emission fields are shown on thesame scale as Fig5with dark blue purple and black pixels denotingemission rates of 25times1016 125times1016 andle 50times1015moleculesper grid square per second respectively The reduction in emissionssince first exposure can be seen with respect to Fig5
being biased towards greater emissions when the seaweedsare first uncovered by the retreating tide the initial burst ofemissions following first exposure is evident in the contribu-tions ofL ochroleucaand particularlySaccharina latissima(which grows in habitats spanning a narrow depth range)
4 I2 measurements during RHaMBLe
41 BBCRDS measurements
A broadband cavity ringdown spectrometer was deployedfrom a shipping container sited on the jetty in front of theSBR adjacent to the containers housing the campaignrsquos other
11828 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 7 Tide height (panel 1) and calculated regional emissions of I2 from 5 to 28 September 2006 (panel 2) The emission contributions aredifferentiated according to seaweed species by colour in panel 2 and are plotted separately in panels 3ndash7
in situ instruments (McFiggans et al 2010) Broadbandcavity ringdown spectroscopy (BBCRDS) uses light from apulsed broadband laser to measure the absorption spectrumof samples contained within a high finesse optical cavity(Bitter et al 2005 Ball and Jones 2003 2009) In this casethe BBCRDS instrument was configured to detect molecu-lar iodine using several of the I2 moleculersquos Blarr X absorp-tion bands in the wavelength range 560ndash570 nm Other at-mospheric gases (H2O NO2 and the oxygen dimer O4) alsoabsorb at these wavelengths and thus contribute to the mea-sured BBCRDS spectra
The BBCRDS system used for this study is based on aninstrument previously used to measure I2 at Mace Head (Ire-land) during the 2002 NAMBLEX campaign (Saiz-Lopezet al 2006 Heard et al 2006) as described in detail byBitteret al (2005) In the intervening years the instrumentrsquos per-formance has been enhanced significantly by upgrading sev-eral key components notably a new laser system that yieldspulsed broadband light with a factor of two wider bandwidth
at green wavelengths a new clocked CCD camera and im-proved analysis softwarespectral fitting routines A broad-band dye laser pumped by a 532 nm NdYAG laser (Sirah Co-bra and Surelight I-20 20 Hz repetition rate) generated lightpulses with an approximately Gaussian emission spectrumcentred at 563 nm (FWHM = 52 nm) This light was directedinto a 187 cm long ringdown cavity formed by two highlyreflective mirrors (Los Gatos peak reflectivity = 99993 at570 nm) Light exiting the ringdown cavity was collected andconveyed through a 100 microm core diameter fibre optic cable toan imaging spectrograph (Chromex 250is) where it was dis-persed in wavelength and imaged onto a clocked CCD cam-era (XCam CCDRem2) The time evolution of individualringdown events was recorded simultaneously at 512 differ-ent wavelengths one for each pixel row of the detector andlight from 50 ringdown events was integrated on the CCDcamera before storing the data to a computer Wavelengthresolved ringdown times were produced by fitting the ring-down decay in each pixel row (j = 1 to 512) The samplersquos
R J Leigh et al Bridging spatial scales in measurements of I2 11829
absorption spectrum was then calculated from sets of ring-down times measured when the cavity contained the sampleτ(λj ) and when flushed with dry nitrogenτ0(λj )
α(λj ) =RL
c
(1
τ(λj )minus
1
τ0(λj )
)=
sumn
αn(λj )+αcon(λj ) (2)
wherec is the speed of light RL is the fraction of the cavitythat is occupied by absorbing speciesαn(λj ) is the wave-length dependent absorption coefficient of the nth molecu-lar absorber andαcon(λj ) is the absorption coefficient dueto all other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16 September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Eq2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken fromVandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported bySaiz-Lopez et al(2004) (see alsoBall et al (2010)) The toppanel of Fig8 shows an example BBCRDS spectrum ob-tained during the campaign where the central and lower pan-els show respectively the I2 and NO2 contributions to themeasured absorption overlaid by their fitted reference spectrafrom the DOAS fitting routine During the RHaMBLe cam-paign the precision of the spectral retrievals was typically10 pptv (parts per trillion by volume) for I2 and 02 ppbv forNO2 (1σ uncertainty 304 s averaging time) Although notthe principal target of this deployment co-retrieval of theNO2 concentrations served as an important quality assuranceparameter with which to monitor the BBCRDS instrumentrsquosperformance Throughout the campaign the NO2 concen-trations measured by BBCRDS were in excellent quantita-tive agreement with NO2 measurements made by the Uni-versity of Yorkrsquos NOxy chemiluminescence instrument as
R J Leigh Bridging spatial scales in measurements of I2 7
α(λj) =RL
c
(1
τ(λj)minus 1τ0(λj)
)=sum
n
αn(λj) + αcon(λj) (2)
where c is the speed of light RL is the fraction of the cav-ity that is occupied by absorbing species αn(λj) is the wave-length dependent absorption coefficient of the nth molecularabsorber and αcon(λj) is the absorption coefficient due toall other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16th September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Equa-tion 2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2
were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken from Vandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported by Saiz-Lopez et al (2004) (see also Ball et al (2010)) The toppanel of Figure 8 shows an example BBCRDS spectrumobtained during the campaign where the central and lowerpanels show respectively the I2 and NO2 contributions tothe measured absorption overlaid by their fitted referencespectra from the DOAS fitting routine During the RHaM-BLe campaign the precision of the spectral retrievals wastypically 10 pptv for I2 and 02 ppbv for NO2 (1σ uncer-tainty 304 s averaging time) Although not the principaltarget of this deployment co-retrieval of the NO2 concentra-tions served as an important quality assurance parameter withwhich to monitor the BBCRDS instrumentrsquos performanceThroughout the campaign the NO2 concentrations measuredby BBCRDS were in excellent quantitative agreement withNO2 measurements made by the University of Yorkrsquos NOxy
chemiluminescence instrument as described in McFigganset al (2010) The NO2 amounts are also a valuable indicator
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14th September 2006 The upper panel shows the measuredspectrum (blue) after subtraction of the absorptions due to watervapour O4 and a second order polynomial accounting for the un-structured absorption contributions The red line shows a DOAS fitto the spectrumrsquos differential structure and the residual spectrumis shown in green The measured (blue) and fitted (red) absorptioncontributions due to I2 and NO2 are shown in the middle and lowerpanels respectively
of the possible extent of I2 recycling via IONO2 chemistryin the semi-polluted environment around Roscoff and so theNO2 field observations from both instruments are shown to-gether in the figures illustrating the measured and modelledI2 concentrations (see Figs 10 12 16 18) The generallygood agreement between the BBCRDS and chemilumines-cence measurements across a wide range of rapidly varyingNO2 concentrations is exemplified by the data from 14th-15th September shown in the bottom panel of Figure 10the gradient of a correlation plot of the NO2 concentrationsrecorded by the two instruments was 098 plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathDifferential Optical Absorption Spectroscopy (LP-DOAS)technique (Plane and Saiz-Lopez 2006) was used to mea-
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14 September 2006 The upper panel shows the measured spec-trum (blue) after subtraction of the absorptions due to water vapourO4 and a second order polynomial accounting for the unstructuredabsorption contributions The red line shows a DOAS fit to the spec-trumrsquos differential structure and the residual spectrum is shown ingreen The measured (blue) and fitted (red) absorption contribu-tions due to I2 and NO2 are shown in the middle and lower panelsrespectively
described inMcFiggans et al(2010) The NO2 amountsare also a valuable indicator of the possible extent of I2recycling via IONO2 chemistry in the semi-polluted envi-ronment around Roscoff and so the NO2 field observationsfrom both instruments are shown together in the figures il-lustrating the measured and modelled I2 concentrations (seeFigs10 12 16 and18) The generally good agreement be-tween the BBCRDS and chemiluminescence measurementsacross a wide range of rapidly varying NO2 concentrations isexemplified by the data from 14ndash15 September shown in thebottom panel of Fig 10 the gradient of a correlation plot ofthe NO2 concentrations recorded by the two instruments was098plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathdifferential optical absorption spectroscopy (LP-DOAS)
11830 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 9 Complete timeseries of model output from 5 to 28 Septem-ber 2006 From top down Regional emissions as calculated bythe model I2 concentrations at the measurement site based on foot-print analysis calculated with photolytic destruction and no recy-cling processes likewise but with recycling at 95 The lowesttwo panels show calculated I2 concentrations along the LP-DOASline of sight with photolytic destruction and no recycling processesand with recycling at 95
technique (Plane and Saiz-Lopez 2006) was used to mea-sure the concentrations of I2 OIO IO and NO3 The absorp-tion path extended 335 km from the SBR (48728 latitudeminus3988 longitude) to a small outcrop on the south west shoreof the Ile de Batz (4874 latitudeminus4036 longitude) wherea retroreflector array was placed to fold the optical path ndash seealso Fig1 The total optical path length was thus 67 km withthe beam 7 to 12 m above the mean sea level Full detailsof the DOAS instrument can be found elsewhere (Mahajanet al 2009 Saiz-Lopez and Plane 2004)
Briefly spectra were recorded with 025 nm resolution be-fore being converted into differential optical density spec-tra The contributions of individual absorbing species to the
Fig 10 Modelled and measured data from 13 to 15 September2006 The top panel shows BBCRDS data (red points and errorbars) with modelled concentrations of I2 at the site assuming 95recycling of I2 photolysed during the daytime (orange line) Themiddle panel shows LP-DOAS data (dark blue points and error bars)with modelled concentrations of I2 in the LP-DOAS light path as-suming 95 I2 recycling (blue line) The grey areas in the uppertwo plots indicate the range of modelled I2 values using recyclingassumptions fromR = 90 to 98 The photolysis frequency of I2is indicated by the green line in the upper two plots and tide by theblack line The bottom panel shows NO2 measured by the NOxychemiluminescence (black) and BBCRDS instruments (red pointsand error bars)
measured spectrum were determined by simultaneous fittingof their molecular absorption cross sections using singularvalue decomposition (Plane and Saiz-Lopez 2006) Aver-aged I2 concentrations along the line of sight were retrievedin the 535minus575 nm window on a number of days and nightsusing the I2 absorption cross sections ofSaiz-Lopez et al(2004) The full data set from the LP-DOAS instrumentis presented inMahajan et al(2009) andMcFiggans et al(2010)
For the present work footprints for the LP-DOAS instru-ment were calculated using the same footprint model (as-suming an 8 m height for the LP-DOAS light beam) withmodelled I2 amounts averaged for the footprints along theline of sight In this way the model provides a path lengthaveraged measurement of I2 along the LP-DOAS light pathwhich sampled emissions from a significant proportion of the
R J Leigh et al Bridging spatial scales in measurements of I2 11831
channel between the SBR and the Ile de Batz (see for exam-ple the footprints in Figs5 and6)
5 Calculation of total emissions and modelled I2 at thesite and along the LP DOAS line of sight
Spatially resolved emissions were calculated for each gridsquare within the model for each one minute time step Siteand LP-DOAS footprints were calculated taking account ofwind speed wind direction and tidal height The time takenfor I2 emissions to travel from their source within the foot-print to the detector was estimated from the windspeed mea-sured at the site In addition to dilution of the emitted I2(accounted for in the footprint) the model also consideredthe photolytic destruction of I2 during its transit to the mea-surement location using equation3
[I2](t)=[I2](0)ej I2middott (3)
where [I2](t) is the volume mixing ratio of I2 at time t and[I2](0) is the volume mixing ratio of I2 at time 0 directlyabove the emission source jI2 is the photolysis frequencyof I2 as measured by a spectral radiometer
Although no chemical modelling was attempted along thelines ofMahajan et al(2009) a simple recycling parameterR was also included in this work to mimic the effects of I2recycling via IONO2 chemistry downstream of the emissionsource This recycling was achieved through modification ofthe I2 photolytic destruction process to
[I2](t)=[I2](0)ej I2middot(1minusR)middott (4)
The recycling parameterR effectively permits a propor-tion of photodissociated I2 to be instantly reformed in ourmodel In order to test this approach the chemical model pre-sented previously inMahajan et al(2009) was re-run in thisstudy to examine the decay of I2 concentrations downwindof an emission source The chemical model was run twiceonce in the presence of NOx (for an assumed baseline NO2concentration of 2 ppbv) and again in the absence of NOxbut for reduced I2 photolysis rates (whilst keeping the pho-tolysis rates of other photolabile species unchanged) Goodqualitative and reasonable quantitative agreement was foundbetween these two scenarios when the I2 photolysis rate wasdecreased to 10 of its typical daytime value providing avalidation for our simplistic approach of reducing the effec-tive I2 photolysis rates to mimic the effect of I2 recycling viaIONO2 formation in this semi-polluted atmosphere For thepresent modelling study the recycling parameter was set to adefault value ofR = 095 ie 95 of the I2 that is photolysedis reformed by subsequent chemistry This is higher than theR = 090 suggested by theMahajan et al(2009) chemicalmodelling but tended to produce the best agreement withthe observational data For comparison modelled I2 concen-trations were also calculated for smaller (R = 090) and larger
(R=098) recycling efficiencies appropriate for respectivelyNO2 concentrations at and above the NO2 = 2 ppbv baselinecase of theMahajan et al(2009) model These results areadditionally shown in Fig 10 and the following other resultsfigures
Figure 9 shows total calculated regional emissions themodelled I2 mixing ratios in air advected to the measurementsite (BBCRDS) and the mean I2 mixing ratio in the air sam-pled along the LP-DOAS line of sight Modelled emissionsin the middle and bottom panels assume 95 recycling ofphotolysed I2 during the daytime Fig9 considers two I2 lossscenarios (i) dilution and irreversible photolytic loss accord-ing to Eq (3) and (ii) dilution and photolytic loss less a re-cycling assumption atR = 095 as per Eq (4) The modelledand measured I2 concentrations are compared in the follow-ing section
6 Comparison of modelled and measured I2
Owing to the challenges inherent in operating the BBCRDSand LP-DOAS instruments in the field and the requirementsfor measurements of other species to be taken by the sameinstrumentation I2 observations are unfortunately not avail-able throughout the campaign Results from three intensivemeasurement periods from each instrument are detailed be-low
61 Comparisons during the night
During the night in the absence of photolytic destructionI2 can be considered as a passive tracer to establish the linkbetween emissions and measurements Agreement betweenthe model and BBCRDS and LP-DOAS measurements atnight indicates acceptable model parameterisation of emis-sion rates seaweed spatial distribution meteorology and di-lution and dispersion within the emission footprint This cantherefore be used as a baseline for investigations during theday when additionally photolytic destruction and chemicalrecycling of I2 become important
Figure10 illustrates night-time measurements and modeldata from 13 to 15 September while Fig11 shows the cor-responding sources of I2 (differentiated by seaweed species)modelled at the measurement site and along the LP-DOASlight path In Fig10 the diurnal cycle is indicated by thephotolysis frequency of I2 (green line) with tide heightsshown in black During the period described in Fig10 therewere night-time low tides shortly after midnight on 14 and15 September 2006 when respectively both the LP-DOASand BBCRDS observed I2 substantially above their detec-tion limits and the model predicted peak I2 concentrations of50ndash100 pptv
Three modelled I2 datasets are presented in Fig10 (andother following data figures) covering the range of I2 re-cycling assumptions discussed in the previous section I2
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
11828 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 7 Tide height (panel 1) and calculated regional emissions of I2 from 5 to 28 September 2006 (panel 2) The emission contributions aredifferentiated according to seaweed species by colour in panel 2 and are plotted separately in panels 3ndash7
in situ instruments (McFiggans et al 2010) Broadbandcavity ringdown spectroscopy (BBCRDS) uses light from apulsed broadband laser to measure the absorption spectrumof samples contained within a high finesse optical cavity(Bitter et al 2005 Ball and Jones 2003 2009) In this casethe BBCRDS instrument was configured to detect molecu-lar iodine using several of the I2 moleculersquos Blarr X absorp-tion bands in the wavelength range 560ndash570 nm Other at-mospheric gases (H2O NO2 and the oxygen dimer O4) alsoabsorb at these wavelengths and thus contribute to the mea-sured BBCRDS spectra
The BBCRDS system used for this study is based on aninstrument previously used to measure I2 at Mace Head (Ire-land) during the 2002 NAMBLEX campaign (Saiz-Lopezet al 2006 Heard et al 2006) as described in detail byBitteret al (2005) In the intervening years the instrumentrsquos per-formance has been enhanced significantly by upgrading sev-eral key components notably a new laser system that yieldspulsed broadband light with a factor of two wider bandwidth
at green wavelengths a new clocked CCD camera and im-proved analysis softwarespectral fitting routines A broad-band dye laser pumped by a 532 nm NdYAG laser (Sirah Co-bra and Surelight I-20 20 Hz repetition rate) generated lightpulses with an approximately Gaussian emission spectrumcentred at 563 nm (FWHM = 52 nm) This light was directedinto a 187 cm long ringdown cavity formed by two highlyreflective mirrors (Los Gatos peak reflectivity = 99993 at570 nm) Light exiting the ringdown cavity was collected andconveyed through a 100 microm core diameter fibre optic cable toan imaging spectrograph (Chromex 250is) where it was dis-persed in wavelength and imaged onto a clocked CCD cam-era (XCam CCDRem2) The time evolution of individualringdown events was recorded simultaneously at 512 differ-ent wavelengths one for each pixel row of the detector andlight from 50 ringdown events was integrated on the CCDcamera before storing the data to a computer Wavelengthresolved ringdown times were produced by fitting the ring-down decay in each pixel row (j = 1 to 512) The samplersquos
R J Leigh et al Bridging spatial scales in measurements of I2 11829
absorption spectrum was then calculated from sets of ring-down times measured when the cavity contained the sampleτ(λj ) and when flushed with dry nitrogenτ0(λj )
α(λj ) =RL
c
(1
τ(λj )minus
1
τ0(λj )
)=
sumn
αn(λj )+αcon(λj ) (2)
wherec is the speed of light RL is the fraction of the cavitythat is occupied by absorbing speciesαn(λj ) is the wave-length dependent absorption coefficient of the nth molecu-lar absorber andαcon(λj ) is the absorption coefficient dueto all other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16 September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Eq2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken fromVandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported bySaiz-Lopez et al(2004) (see alsoBall et al (2010)) The toppanel of Fig8 shows an example BBCRDS spectrum ob-tained during the campaign where the central and lower pan-els show respectively the I2 and NO2 contributions to themeasured absorption overlaid by their fitted reference spectrafrom the DOAS fitting routine During the RHaMBLe cam-paign the precision of the spectral retrievals was typically10 pptv (parts per trillion by volume) for I2 and 02 ppbv forNO2 (1σ uncertainty 304 s averaging time) Although notthe principal target of this deployment co-retrieval of theNO2 concentrations served as an important quality assuranceparameter with which to monitor the BBCRDS instrumentrsquosperformance Throughout the campaign the NO2 concen-trations measured by BBCRDS were in excellent quantita-tive agreement with NO2 measurements made by the Uni-versity of Yorkrsquos NOxy chemiluminescence instrument as
R J Leigh Bridging spatial scales in measurements of I2 7
α(λj) =RL
c
(1
τ(λj)minus 1τ0(λj)
)=sum
n
αn(λj) + αcon(λj) (2)
where c is the speed of light RL is the fraction of the cav-ity that is occupied by absorbing species αn(λj) is the wave-length dependent absorption coefficient of the nth molecularabsorber and αcon(λj) is the absorption coefficient due toall other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16th September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Equa-tion 2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2
were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken from Vandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported by Saiz-Lopez et al (2004) (see also Ball et al (2010)) The toppanel of Figure 8 shows an example BBCRDS spectrumobtained during the campaign where the central and lowerpanels show respectively the I2 and NO2 contributions tothe measured absorption overlaid by their fitted referencespectra from the DOAS fitting routine During the RHaM-BLe campaign the precision of the spectral retrievals wastypically 10 pptv for I2 and 02 ppbv for NO2 (1σ uncer-tainty 304 s averaging time) Although not the principaltarget of this deployment co-retrieval of the NO2 concentra-tions served as an important quality assurance parameter withwhich to monitor the BBCRDS instrumentrsquos performanceThroughout the campaign the NO2 concentrations measuredby BBCRDS were in excellent quantitative agreement withNO2 measurements made by the University of Yorkrsquos NOxy
chemiluminescence instrument as described in McFigganset al (2010) The NO2 amounts are also a valuable indicator
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14th September 2006 The upper panel shows the measuredspectrum (blue) after subtraction of the absorptions due to watervapour O4 and a second order polynomial accounting for the un-structured absorption contributions The red line shows a DOAS fitto the spectrumrsquos differential structure and the residual spectrumis shown in green The measured (blue) and fitted (red) absorptioncontributions due to I2 and NO2 are shown in the middle and lowerpanels respectively
of the possible extent of I2 recycling via IONO2 chemistryin the semi-polluted environment around Roscoff and so theNO2 field observations from both instruments are shown to-gether in the figures illustrating the measured and modelledI2 concentrations (see Figs 10 12 16 18) The generallygood agreement between the BBCRDS and chemilumines-cence measurements across a wide range of rapidly varyingNO2 concentrations is exemplified by the data from 14th-15th September shown in the bottom panel of Figure 10the gradient of a correlation plot of the NO2 concentrationsrecorded by the two instruments was 098 plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathDifferential Optical Absorption Spectroscopy (LP-DOAS)technique (Plane and Saiz-Lopez 2006) was used to mea-
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14 September 2006 The upper panel shows the measured spec-trum (blue) after subtraction of the absorptions due to water vapourO4 and a second order polynomial accounting for the unstructuredabsorption contributions The red line shows a DOAS fit to the spec-trumrsquos differential structure and the residual spectrum is shown ingreen The measured (blue) and fitted (red) absorption contribu-tions due to I2 and NO2 are shown in the middle and lower panelsrespectively
described inMcFiggans et al(2010) The NO2 amountsare also a valuable indicator of the possible extent of I2recycling via IONO2 chemistry in the semi-polluted envi-ronment around Roscoff and so the NO2 field observationsfrom both instruments are shown together in the figures il-lustrating the measured and modelled I2 concentrations (seeFigs10 12 16 and18) The generally good agreement be-tween the BBCRDS and chemiluminescence measurementsacross a wide range of rapidly varying NO2 concentrations isexemplified by the data from 14ndash15 September shown in thebottom panel of Fig 10 the gradient of a correlation plot ofthe NO2 concentrations recorded by the two instruments was098plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathdifferential optical absorption spectroscopy (LP-DOAS)
11830 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 9 Complete timeseries of model output from 5 to 28 Septem-ber 2006 From top down Regional emissions as calculated bythe model I2 concentrations at the measurement site based on foot-print analysis calculated with photolytic destruction and no recy-cling processes likewise but with recycling at 95 The lowesttwo panels show calculated I2 concentrations along the LP-DOASline of sight with photolytic destruction and no recycling processesand with recycling at 95
technique (Plane and Saiz-Lopez 2006) was used to mea-sure the concentrations of I2 OIO IO and NO3 The absorp-tion path extended 335 km from the SBR (48728 latitudeminus3988 longitude) to a small outcrop on the south west shoreof the Ile de Batz (4874 latitudeminus4036 longitude) wherea retroreflector array was placed to fold the optical path ndash seealso Fig1 The total optical path length was thus 67 km withthe beam 7 to 12 m above the mean sea level Full detailsof the DOAS instrument can be found elsewhere (Mahajanet al 2009 Saiz-Lopez and Plane 2004)
Briefly spectra were recorded with 025 nm resolution be-fore being converted into differential optical density spec-tra The contributions of individual absorbing species to the
Fig 10 Modelled and measured data from 13 to 15 September2006 The top panel shows BBCRDS data (red points and errorbars) with modelled concentrations of I2 at the site assuming 95recycling of I2 photolysed during the daytime (orange line) Themiddle panel shows LP-DOAS data (dark blue points and error bars)with modelled concentrations of I2 in the LP-DOAS light path as-suming 95 I2 recycling (blue line) The grey areas in the uppertwo plots indicate the range of modelled I2 values using recyclingassumptions fromR = 90 to 98 The photolysis frequency of I2is indicated by the green line in the upper two plots and tide by theblack line The bottom panel shows NO2 measured by the NOxychemiluminescence (black) and BBCRDS instruments (red pointsand error bars)
measured spectrum were determined by simultaneous fittingof their molecular absorption cross sections using singularvalue decomposition (Plane and Saiz-Lopez 2006) Aver-aged I2 concentrations along the line of sight were retrievedin the 535minus575 nm window on a number of days and nightsusing the I2 absorption cross sections ofSaiz-Lopez et al(2004) The full data set from the LP-DOAS instrumentis presented inMahajan et al(2009) andMcFiggans et al(2010)
For the present work footprints for the LP-DOAS instru-ment were calculated using the same footprint model (as-suming an 8 m height for the LP-DOAS light beam) withmodelled I2 amounts averaged for the footprints along theline of sight In this way the model provides a path lengthaveraged measurement of I2 along the LP-DOAS light pathwhich sampled emissions from a significant proportion of the
R J Leigh et al Bridging spatial scales in measurements of I2 11831
channel between the SBR and the Ile de Batz (see for exam-ple the footprints in Figs5 and6)
5 Calculation of total emissions and modelled I2 at thesite and along the LP DOAS line of sight
Spatially resolved emissions were calculated for each gridsquare within the model for each one minute time step Siteand LP-DOAS footprints were calculated taking account ofwind speed wind direction and tidal height The time takenfor I2 emissions to travel from their source within the foot-print to the detector was estimated from the windspeed mea-sured at the site In addition to dilution of the emitted I2(accounted for in the footprint) the model also consideredthe photolytic destruction of I2 during its transit to the mea-surement location using equation3
[I2](t)=[I2](0)ej I2middott (3)
where [I2](t) is the volume mixing ratio of I2 at time t and[I2](0) is the volume mixing ratio of I2 at time 0 directlyabove the emission source jI2 is the photolysis frequencyof I2 as measured by a spectral radiometer
Although no chemical modelling was attempted along thelines ofMahajan et al(2009) a simple recycling parameterR was also included in this work to mimic the effects of I2recycling via IONO2 chemistry downstream of the emissionsource This recycling was achieved through modification ofthe I2 photolytic destruction process to
[I2](t)=[I2](0)ej I2middot(1minusR)middott (4)
The recycling parameterR effectively permits a propor-tion of photodissociated I2 to be instantly reformed in ourmodel In order to test this approach the chemical model pre-sented previously inMahajan et al(2009) was re-run in thisstudy to examine the decay of I2 concentrations downwindof an emission source The chemical model was run twiceonce in the presence of NOx (for an assumed baseline NO2concentration of 2 ppbv) and again in the absence of NOxbut for reduced I2 photolysis rates (whilst keeping the pho-tolysis rates of other photolabile species unchanged) Goodqualitative and reasonable quantitative agreement was foundbetween these two scenarios when the I2 photolysis rate wasdecreased to 10 of its typical daytime value providing avalidation for our simplistic approach of reducing the effec-tive I2 photolysis rates to mimic the effect of I2 recycling viaIONO2 formation in this semi-polluted atmosphere For thepresent modelling study the recycling parameter was set to adefault value ofR = 095 ie 95 of the I2 that is photolysedis reformed by subsequent chemistry This is higher than theR = 090 suggested by theMahajan et al(2009) chemicalmodelling but tended to produce the best agreement withthe observational data For comparison modelled I2 concen-trations were also calculated for smaller (R = 090) and larger
(R=098) recycling efficiencies appropriate for respectivelyNO2 concentrations at and above the NO2 = 2 ppbv baselinecase of theMahajan et al(2009) model These results areadditionally shown in Fig 10 and the following other resultsfigures
Figure 9 shows total calculated regional emissions themodelled I2 mixing ratios in air advected to the measurementsite (BBCRDS) and the mean I2 mixing ratio in the air sam-pled along the LP-DOAS line of sight Modelled emissionsin the middle and bottom panels assume 95 recycling ofphotolysed I2 during the daytime Fig9 considers two I2 lossscenarios (i) dilution and irreversible photolytic loss accord-ing to Eq (3) and (ii) dilution and photolytic loss less a re-cycling assumption atR = 095 as per Eq (4) The modelledand measured I2 concentrations are compared in the follow-ing section
6 Comparison of modelled and measured I2
Owing to the challenges inherent in operating the BBCRDSand LP-DOAS instruments in the field and the requirementsfor measurements of other species to be taken by the sameinstrumentation I2 observations are unfortunately not avail-able throughout the campaign Results from three intensivemeasurement periods from each instrument are detailed be-low
61 Comparisons during the night
During the night in the absence of photolytic destructionI2 can be considered as a passive tracer to establish the linkbetween emissions and measurements Agreement betweenthe model and BBCRDS and LP-DOAS measurements atnight indicates acceptable model parameterisation of emis-sion rates seaweed spatial distribution meteorology and di-lution and dispersion within the emission footprint This cantherefore be used as a baseline for investigations during theday when additionally photolytic destruction and chemicalrecycling of I2 become important
Figure10 illustrates night-time measurements and modeldata from 13 to 15 September while Fig11 shows the cor-responding sources of I2 (differentiated by seaweed species)modelled at the measurement site and along the LP-DOASlight path In Fig10 the diurnal cycle is indicated by thephotolysis frequency of I2 (green line) with tide heightsshown in black During the period described in Fig10 therewere night-time low tides shortly after midnight on 14 and15 September 2006 when respectively both the LP-DOASand BBCRDS observed I2 substantially above their detec-tion limits and the model predicted peak I2 concentrations of50ndash100 pptv
Three modelled I2 datasets are presented in Fig10 (andother following data figures) covering the range of I2 re-cycling assumptions discussed in the previous section I2
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
R J Leigh et al Bridging spatial scales in measurements of I2 11829
absorption spectrum was then calculated from sets of ring-down times measured when the cavity contained the sampleτ(λj ) and when flushed with dry nitrogenτ0(λj )
α(λj ) =RL
c
(1
τ(λj )minus
1
τ0(λj )
)=
sumn
αn(λj )+αcon(λj ) (2)
wherec is the speed of light RL is the fraction of the cavitythat is occupied by absorbing speciesαn(λj ) is the wave-length dependent absorption coefficient of the nth molecu-lar absorber andαcon(λj ) is the absorption coefficient dueto all other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16 September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Eq2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken fromVandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported bySaiz-Lopez et al(2004) (see alsoBall et al (2010)) The toppanel of Fig8 shows an example BBCRDS spectrum ob-tained during the campaign where the central and lower pan-els show respectively the I2 and NO2 contributions to themeasured absorption overlaid by their fitted reference spectrafrom the DOAS fitting routine During the RHaMBLe cam-paign the precision of the spectral retrievals was typically10 pptv (parts per trillion by volume) for I2 and 02 ppbv forNO2 (1σ uncertainty 304 s averaging time) Although notthe principal target of this deployment co-retrieval of theNO2 concentrations served as an important quality assuranceparameter with which to monitor the BBCRDS instrumentrsquosperformance Throughout the campaign the NO2 concen-trations measured by BBCRDS were in excellent quantita-tive agreement with NO2 measurements made by the Uni-versity of Yorkrsquos NOxy chemiluminescence instrument as
R J Leigh Bridging spatial scales in measurements of I2 7
α(λj) =RL
c
(1
τ(λj)minus 1τ0(λj)
)=sum
n
αn(λj) + αcon(λj) (2)
where c is the speed of light RL is the fraction of the cav-ity that is occupied by absorbing species αn(λj) is the wave-length dependent absorption coefficient of the nth molecularabsorber and αcon(λj) is the absorption coefficient due toall other contributions to the spectrumrsquos unstructured con-tinuum absorption (mainly aerosol extinction) During thefirst part of the campaign (before 16th September) the cav-ity was located inside the shipping container and ambient airwas drawn into the cavity at 3 litres per minute The cav-ity was then moved onto the roof of the container and oper-ated in an open-path configuration for the remainder of thecampaign In both cases appropriate corrections (Shillings2009) were made to account for exclusion of the atmosphericsample from the cavityrsquos mirror mounts which were purgedwith dry nitrogen to prevent contamination of the optical sur-faces by ambient aerosol particles (ie the RL term in Equa-tion 2)
BBCRDS absorption spectra were averaged to a time reso-lution of 5 minutes and the known absorptions due to ambientH2O (humidity meter) and O4 (atmospheric oxygen concen-tration) were subtracted The concentrations of I2 and NO2
were then retrieved from a multivariate fit of reference ab-sorption cross sections to the structured features remainingin the samplersquos absorption spectrum using an analysis simi-lar to that developed for DOAS (Platt 1999 Ball and Jones2003 2009) NO2 cross sections were taken from Vandaeleet al (1996) and were degraded to the 012 nm FWHM in-strumental resolution I2 cross sections were derived fromthe PGOPHER spectral simulation program (Western Ac-cess September 2009 Martin et al 1986) and were scaledto reproduce the differential cross sections reported by Saiz-Lopez et al (2004) (see also Ball et al (2010)) The toppanel of Figure 8 shows an example BBCRDS spectrumobtained during the campaign where the central and lowerpanels show respectively the I2 and NO2 contributions tothe measured absorption overlaid by their fitted referencespectra from the DOAS fitting routine During the RHaM-BLe campaign the precision of the spectral retrievals wastypically 10 pptv for I2 and 02 ppbv for NO2 (1σ uncer-tainty 304 s averaging time) Although not the principaltarget of this deployment co-retrieval of the NO2 concentra-tions served as an important quality assurance parameter withwhich to monitor the BBCRDS instrumentrsquos performanceThroughout the campaign the NO2 concentrations measuredby BBCRDS were in excellent quantitative agreement withNO2 measurements made by the University of Yorkrsquos NOxy
chemiluminescence instrument as described in McFigganset al (2010) The NO2 amounts are also a valuable indicator
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14th September 2006 The upper panel shows the measuredspectrum (blue) after subtraction of the absorptions due to watervapour O4 and a second order polynomial accounting for the un-structured absorption contributions The red line shows a DOAS fitto the spectrumrsquos differential structure and the residual spectrumis shown in green The measured (blue) and fitted (red) absorptioncontributions due to I2 and NO2 are shown in the middle and lowerpanels respectively
of the possible extent of I2 recycling via IONO2 chemistryin the semi-polluted environment around Roscoff and so theNO2 field observations from both instruments are shown to-gether in the figures illustrating the measured and modelledI2 concentrations (see Figs 10 12 16 18) The generallygood agreement between the BBCRDS and chemilumines-cence measurements across a wide range of rapidly varyingNO2 concentrations is exemplified by the data from 14th-15th September shown in the bottom panel of Figure 10the gradient of a correlation plot of the NO2 concentrationsrecorded by the two instruments was 098 plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathDifferential Optical Absorption Spectroscopy (LP-DOAS)technique (Plane and Saiz-Lopez 2006) was used to mea-
Fig 8 An example BBCRDS spectrum recorded around 1730 UTon 14 September 2006 The upper panel shows the measured spec-trum (blue) after subtraction of the absorptions due to water vapourO4 and a second order polynomial accounting for the unstructuredabsorption contributions The red line shows a DOAS fit to the spec-trumrsquos differential structure and the residual spectrum is shown ingreen The measured (blue) and fitted (red) absorption contribu-tions due to I2 and NO2 are shown in the middle and lower panelsrespectively
described inMcFiggans et al(2010) The NO2 amountsare also a valuable indicator of the possible extent of I2recycling via IONO2 chemistry in the semi-polluted envi-ronment around Roscoff and so the NO2 field observationsfrom both instruments are shown together in the figures il-lustrating the measured and modelled I2 concentrations (seeFigs10 12 16 and18) The generally good agreement be-tween the BBCRDS and chemiluminescence measurementsacross a wide range of rapidly varying NO2 concentrations isexemplified by the data from 14ndash15 September shown in thebottom panel of Fig 10 the gradient of a correlation plot ofthe NO2 concentrations recorded by the two instruments was098plusmn 003
42 Measurements taken by long path DOAS
During the RHaMBLe Roscoff deployment the long pathdifferential optical absorption spectroscopy (LP-DOAS)
11830 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 9 Complete timeseries of model output from 5 to 28 Septem-ber 2006 From top down Regional emissions as calculated bythe model I2 concentrations at the measurement site based on foot-print analysis calculated with photolytic destruction and no recy-cling processes likewise but with recycling at 95 The lowesttwo panels show calculated I2 concentrations along the LP-DOASline of sight with photolytic destruction and no recycling processesand with recycling at 95
technique (Plane and Saiz-Lopez 2006) was used to mea-sure the concentrations of I2 OIO IO and NO3 The absorp-tion path extended 335 km from the SBR (48728 latitudeminus3988 longitude) to a small outcrop on the south west shoreof the Ile de Batz (4874 latitudeminus4036 longitude) wherea retroreflector array was placed to fold the optical path ndash seealso Fig1 The total optical path length was thus 67 km withthe beam 7 to 12 m above the mean sea level Full detailsof the DOAS instrument can be found elsewhere (Mahajanet al 2009 Saiz-Lopez and Plane 2004)
Briefly spectra were recorded with 025 nm resolution be-fore being converted into differential optical density spec-tra The contributions of individual absorbing species to the
Fig 10 Modelled and measured data from 13 to 15 September2006 The top panel shows BBCRDS data (red points and errorbars) with modelled concentrations of I2 at the site assuming 95recycling of I2 photolysed during the daytime (orange line) Themiddle panel shows LP-DOAS data (dark blue points and error bars)with modelled concentrations of I2 in the LP-DOAS light path as-suming 95 I2 recycling (blue line) The grey areas in the uppertwo plots indicate the range of modelled I2 values using recyclingassumptions fromR = 90 to 98 The photolysis frequency of I2is indicated by the green line in the upper two plots and tide by theblack line The bottom panel shows NO2 measured by the NOxychemiluminescence (black) and BBCRDS instruments (red pointsand error bars)
measured spectrum were determined by simultaneous fittingof their molecular absorption cross sections using singularvalue decomposition (Plane and Saiz-Lopez 2006) Aver-aged I2 concentrations along the line of sight were retrievedin the 535minus575 nm window on a number of days and nightsusing the I2 absorption cross sections ofSaiz-Lopez et al(2004) The full data set from the LP-DOAS instrumentis presented inMahajan et al(2009) andMcFiggans et al(2010)
For the present work footprints for the LP-DOAS instru-ment were calculated using the same footprint model (as-suming an 8 m height for the LP-DOAS light beam) withmodelled I2 amounts averaged for the footprints along theline of sight In this way the model provides a path lengthaveraged measurement of I2 along the LP-DOAS light pathwhich sampled emissions from a significant proportion of the
R J Leigh et al Bridging spatial scales in measurements of I2 11831
channel between the SBR and the Ile de Batz (see for exam-ple the footprints in Figs5 and6)
5 Calculation of total emissions and modelled I2 at thesite and along the LP DOAS line of sight
Spatially resolved emissions were calculated for each gridsquare within the model for each one minute time step Siteand LP-DOAS footprints were calculated taking account ofwind speed wind direction and tidal height The time takenfor I2 emissions to travel from their source within the foot-print to the detector was estimated from the windspeed mea-sured at the site In addition to dilution of the emitted I2(accounted for in the footprint) the model also consideredthe photolytic destruction of I2 during its transit to the mea-surement location using equation3
[I2](t)=[I2](0)ej I2middott (3)
where [I2](t) is the volume mixing ratio of I2 at time t and[I2](0) is the volume mixing ratio of I2 at time 0 directlyabove the emission source jI2 is the photolysis frequencyof I2 as measured by a spectral radiometer
Although no chemical modelling was attempted along thelines ofMahajan et al(2009) a simple recycling parameterR was also included in this work to mimic the effects of I2recycling via IONO2 chemistry downstream of the emissionsource This recycling was achieved through modification ofthe I2 photolytic destruction process to
[I2](t)=[I2](0)ej I2middot(1minusR)middott (4)
The recycling parameterR effectively permits a propor-tion of photodissociated I2 to be instantly reformed in ourmodel In order to test this approach the chemical model pre-sented previously inMahajan et al(2009) was re-run in thisstudy to examine the decay of I2 concentrations downwindof an emission source The chemical model was run twiceonce in the presence of NOx (for an assumed baseline NO2concentration of 2 ppbv) and again in the absence of NOxbut for reduced I2 photolysis rates (whilst keeping the pho-tolysis rates of other photolabile species unchanged) Goodqualitative and reasonable quantitative agreement was foundbetween these two scenarios when the I2 photolysis rate wasdecreased to 10 of its typical daytime value providing avalidation for our simplistic approach of reducing the effec-tive I2 photolysis rates to mimic the effect of I2 recycling viaIONO2 formation in this semi-polluted atmosphere For thepresent modelling study the recycling parameter was set to adefault value ofR = 095 ie 95 of the I2 that is photolysedis reformed by subsequent chemistry This is higher than theR = 090 suggested by theMahajan et al(2009) chemicalmodelling but tended to produce the best agreement withthe observational data For comparison modelled I2 concen-trations were also calculated for smaller (R = 090) and larger
(R=098) recycling efficiencies appropriate for respectivelyNO2 concentrations at and above the NO2 = 2 ppbv baselinecase of theMahajan et al(2009) model These results areadditionally shown in Fig 10 and the following other resultsfigures
Figure 9 shows total calculated regional emissions themodelled I2 mixing ratios in air advected to the measurementsite (BBCRDS) and the mean I2 mixing ratio in the air sam-pled along the LP-DOAS line of sight Modelled emissionsin the middle and bottom panels assume 95 recycling ofphotolysed I2 during the daytime Fig9 considers two I2 lossscenarios (i) dilution and irreversible photolytic loss accord-ing to Eq (3) and (ii) dilution and photolytic loss less a re-cycling assumption atR = 095 as per Eq (4) The modelledand measured I2 concentrations are compared in the follow-ing section
6 Comparison of modelled and measured I2
Owing to the challenges inherent in operating the BBCRDSand LP-DOAS instruments in the field and the requirementsfor measurements of other species to be taken by the sameinstrumentation I2 observations are unfortunately not avail-able throughout the campaign Results from three intensivemeasurement periods from each instrument are detailed be-low
61 Comparisons during the night
During the night in the absence of photolytic destructionI2 can be considered as a passive tracer to establish the linkbetween emissions and measurements Agreement betweenthe model and BBCRDS and LP-DOAS measurements atnight indicates acceptable model parameterisation of emis-sion rates seaweed spatial distribution meteorology and di-lution and dispersion within the emission footprint This cantherefore be used as a baseline for investigations during theday when additionally photolytic destruction and chemicalrecycling of I2 become important
Figure10 illustrates night-time measurements and modeldata from 13 to 15 September while Fig11 shows the cor-responding sources of I2 (differentiated by seaweed species)modelled at the measurement site and along the LP-DOASlight path In Fig10 the diurnal cycle is indicated by thephotolysis frequency of I2 (green line) with tide heightsshown in black During the period described in Fig10 therewere night-time low tides shortly after midnight on 14 and15 September 2006 when respectively both the LP-DOASand BBCRDS observed I2 substantially above their detec-tion limits and the model predicted peak I2 concentrations of50ndash100 pptv
Three modelled I2 datasets are presented in Fig10 (andother following data figures) covering the range of I2 re-cycling assumptions discussed in the previous section I2
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
11830 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 9 Complete timeseries of model output from 5 to 28 Septem-ber 2006 From top down Regional emissions as calculated bythe model I2 concentrations at the measurement site based on foot-print analysis calculated with photolytic destruction and no recy-cling processes likewise but with recycling at 95 The lowesttwo panels show calculated I2 concentrations along the LP-DOASline of sight with photolytic destruction and no recycling processesand with recycling at 95
technique (Plane and Saiz-Lopez 2006) was used to mea-sure the concentrations of I2 OIO IO and NO3 The absorp-tion path extended 335 km from the SBR (48728 latitudeminus3988 longitude) to a small outcrop on the south west shoreof the Ile de Batz (4874 latitudeminus4036 longitude) wherea retroreflector array was placed to fold the optical path ndash seealso Fig1 The total optical path length was thus 67 km withthe beam 7 to 12 m above the mean sea level Full detailsof the DOAS instrument can be found elsewhere (Mahajanet al 2009 Saiz-Lopez and Plane 2004)
Briefly spectra were recorded with 025 nm resolution be-fore being converted into differential optical density spec-tra The contributions of individual absorbing species to the
Fig 10 Modelled and measured data from 13 to 15 September2006 The top panel shows BBCRDS data (red points and errorbars) with modelled concentrations of I2 at the site assuming 95recycling of I2 photolysed during the daytime (orange line) Themiddle panel shows LP-DOAS data (dark blue points and error bars)with modelled concentrations of I2 in the LP-DOAS light path as-suming 95 I2 recycling (blue line) The grey areas in the uppertwo plots indicate the range of modelled I2 values using recyclingassumptions fromR = 90 to 98 The photolysis frequency of I2is indicated by the green line in the upper two plots and tide by theblack line The bottom panel shows NO2 measured by the NOxychemiluminescence (black) and BBCRDS instruments (red pointsand error bars)
measured spectrum were determined by simultaneous fittingof their molecular absorption cross sections using singularvalue decomposition (Plane and Saiz-Lopez 2006) Aver-aged I2 concentrations along the line of sight were retrievedin the 535minus575 nm window on a number of days and nightsusing the I2 absorption cross sections ofSaiz-Lopez et al(2004) The full data set from the LP-DOAS instrumentis presented inMahajan et al(2009) andMcFiggans et al(2010)
For the present work footprints for the LP-DOAS instru-ment were calculated using the same footprint model (as-suming an 8 m height for the LP-DOAS light beam) withmodelled I2 amounts averaged for the footprints along theline of sight In this way the model provides a path lengthaveraged measurement of I2 along the LP-DOAS light pathwhich sampled emissions from a significant proportion of the
R J Leigh et al Bridging spatial scales in measurements of I2 11831
channel between the SBR and the Ile de Batz (see for exam-ple the footprints in Figs5 and6)
5 Calculation of total emissions and modelled I2 at thesite and along the LP DOAS line of sight
Spatially resolved emissions were calculated for each gridsquare within the model for each one minute time step Siteand LP-DOAS footprints were calculated taking account ofwind speed wind direction and tidal height The time takenfor I2 emissions to travel from their source within the foot-print to the detector was estimated from the windspeed mea-sured at the site In addition to dilution of the emitted I2(accounted for in the footprint) the model also consideredthe photolytic destruction of I2 during its transit to the mea-surement location using equation3
[I2](t)=[I2](0)ej I2middott (3)
where [I2](t) is the volume mixing ratio of I2 at time t and[I2](0) is the volume mixing ratio of I2 at time 0 directlyabove the emission source jI2 is the photolysis frequencyof I2 as measured by a spectral radiometer
Although no chemical modelling was attempted along thelines ofMahajan et al(2009) a simple recycling parameterR was also included in this work to mimic the effects of I2recycling via IONO2 chemistry downstream of the emissionsource This recycling was achieved through modification ofthe I2 photolytic destruction process to
[I2](t)=[I2](0)ej I2middot(1minusR)middott (4)
The recycling parameterR effectively permits a propor-tion of photodissociated I2 to be instantly reformed in ourmodel In order to test this approach the chemical model pre-sented previously inMahajan et al(2009) was re-run in thisstudy to examine the decay of I2 concentrations downwindof an emission source The chemical model was run twiceonce in the presence of NOx (for an assumed baseline NO2concentration of 2 ppbv) and again in the absence of NOxbut for reduced I2 photolysis rates (whilst keeping the pho-tolysis rates of other photolabile species unchanged) Goodqualitative and reasonable quantitative agreement was foundbetween these two scenarios when the I2 photolysis rate wasdecreased to 10 of its typical daytime value providing avalidation for our simplistic approach of reducing the effec-tive I2 photolysis rates to mimic the effect of I2 recycling viaIONO2 formation in this semi-polluted atmosphere For thepresent modelling study the recycling parameter was set to adefault value ofR = 095 ie 95 of the I2 that is photolysedis reformed by subsequent chemistry This is higher than theR = 090 suggested by theMahajan et al(2009) chemicalmodelling but tended to produce the best agreement withthe observational data For comparison modelled I2 concen-trations were also calculated for smaller (R = 090) and larger
(R=098) recycling efficiencies appropriate for respectivelyNO2 concentrations at and above the NO2 = 2 ppbv baselinecase of theMahajan et al(2009) model These results areadditionally shown in Fig 10 and the following other resultsfigures
Figure 9 shows total calculated regional emissions themodelled I2 mixing ratios in air advected to the measurementsite (BBCRDS) and the mean I2 mixing ratio in the air sam-pled along the LP-DOAS line of sight Modelled emissionsin the middle and bottom panels assume 95 recycling ofphotolysed I2 during the daytime Fig9 considers two I2 lossscenarios (i) dilution and irreversible photolytic loss accord-ing to Eq (3) and (ii) dilution and photolytic loss less a re-cycling assumption atR = 095 as per Eq (4) The modelledand measured I2 concentrations are compared in the follow-ing section
6 Comparison of modelled and measured I2
Owing to the challenges inherent in operating the BBCRDSand LP-DOAS instruments in the field and the requirementsfor measurements of other species to be taken by the sameinstrumentation I2 observations are unfortunately not avail-able throughout the campaign Results from three intensivemeasurement periods from each instrument are detailed be-low
61 Comparisons during the night
During the night in the absence of photolytic destructionI2 can be considered as a passive tracer to establish the linkbetween emissions and measurements Agreement betweenthe model and BBCRDS and LP-DOAS measurements atnight indicates acceptable model parameterisation of emis-sion rates seaweed spatial distribution meteorology and di-lution and dispersion within the emission footprint This cantherefore be used as a baseline for investigations during theday when additionally photolytic destruction and chemicalrecycling of I2 become important
Figure10 illustrates night-time measurements and modeldata from 13 to 15 September while Fig11 shows the cor-responding sources of I2 (differentiated by seaweed species)modelled at the measurement site and along the LP-DOASlight path In Fig10 the diurnal cycle is indicated by thephotolysis frequency of I2 (green line) with tide heightsshown in black During the period described in Fig10 therewere night-time low tides shortly after midnight on 14 and15 September 2006 when respectively both the LP-DOASand BBCRDS observed I2 substantially above their detec-tion limits and the model predicted peak I2 concentrations of50ndash100 pptv
Three modelled I2 datasets are presented in Fig10 (andother following data figures) covering the range of I2 re-cycling assumptions discussed in the previous section I2
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
R J Leigh et al Bridging spatial scales in measurements of I2 11831
channel between the SBR and the Ile de Batz (see for exam-ple the footprints in Figs5 and6)
5 Calculation of total emissions and modelled I2 at thesite and along the LP DOAS line of sight
Spatially resolved emissions were calculated for each gridsquare within the model for each one minute time step Siteand LP-DOAS footprints were calculated taking account ofwind speed wind direction and tidal height The time takenfor I2 emissions to travel from their source within the foot-print to the detector was estimated from the windspeed mea-sured at the site In addition to dilution of the emitted I2(accounted for in the footprint) the model also consideredthe photolytic destruction of I2 during its transit to the mea-surement location using equation3
[I2](t)=[I2](0)ej I2middott (3)
where [I2](t) is the volume mixing ratio of I2 at time t and[I2](0) is the volume mixing ratio of I2 at time 0 directlyabove the emission source jI2 is the photolysis frequencyof I2 as measured by a spectral radiometer
Although no chemical modelling was attempted along thelines ofMahajan et al(2009) a simple recycling parameterR was also included in this work to mimic the effects of I2recycling via IONO2 chemistry downstream of the emissionsource This recycling was achieved through modification ofthe I2 photolytic destruction process to
[I2](t)=[I2](0)ej I2middot(1minusR)middott (4)
The recycling parameterR effectively permits a propor-tion of photodissociated I2 to be instantly reformed in ourmodel In order to test this approach the chemical model pre-sented previously inMahajan et al(2009) was re-run in thisstudy to examine the decay of I2 concentrations downwindof an emission source The chemical model was run twiceonce in the presence of NOx (for an assumed baseline NO2concentration of 2 ppbv) and again in the absence of NOxbut for reduced I2 photolysis rates (whilst keeping the pho-tolysis rates of other photolabile species unchanged) Goodqualitative and reasonable quantitative agreement was foundbetween these two scenarios when the I2 photolysis rate wasdecreased to 10 of its typical daytime value providing avalidation for our simplistic approach of reducing the effec-tive I2 photolysis rates to mimic the effect of I2 recycling viaIONO2 formation in this semi-polluted atmosphere For thepresent modelling study the recycling parameter was set to adefault value ofR = 095 ie 95 of the I2 that is photolysedis reformed by subsequent chemistry This is higher than theR = 090 suggested by theMahajan et al(2009) chemicalmodelling but tended to produce the best agreement withthe observational data For comparison modelled I2 concen-trations were also calculated for smaller (R = 090) and larger
(R=098) recycling efficiencies appropriate for respectivelyNO2 concentrations at and above the NO2 = 2 ppbv baselinecase of theMahajan et al(2009) model These results areadditionally shown in Fig 10 and the following other resultsfigures
Figure 9 shows total calculated regional emissions themodelled I2 mixing ratios in air advected to the measurementsite (BBCRDS) and the mean I2 mixing ratio in the air sam-pled along the LP-DOAS line of sight Modelled emissionsin the middle and bottom panels assume 95 recycling ofphotolysed I2 during the daytime Fig9 considers two I2 lossscenarios (i) dilution and irreversible photolytic loss accord-ing to Eq (3) and (ii) dilution and photolytic loss less a re-cycling assumption atR = 095 as per Eq (4) The modelledand measured I2 concentrations are compared in the follow-ing section
6 Comparison of modelled and measured I2
Owing to the challenges inherent in operating the BBCRDSand LP-DOAS instruments in the field and the requirementsfor measurements of other species to be taken by the sameinstrumentation I2 observations are unfortunately not avail-able throughout the campaign Results from three intensivemeasurement periods from each instrument are detailed be-low
61 Comparisons during the night
During the night in the absence of photolytic destructionI2 can be considered as a passive tracer to establish the linkbetween emissions and measurements Agreement betweenthe model and BBCRDS and LP-DOAS measurements atnight indicates acceptable model parameterisation of emis-sion rates seaweed spatial distribution meteorology and di-lution and dispersion within the emission footprint This cantherefore be used as a baseline for investigations during theday when additionally photolytic destruction and chemicalrecycling of I2 become important
Figure10 illustrates night-time measurements and modeldata from 13 to 15 September while Fig11 shows the cor-responding sources of I2 (differentiated by seaweed species)modelled at the measurement site and along the LP-DOASlight path In Fig10 the diurnal cycle is indicated by thephotolysis frequency of I2 (green line) with tide heightsshown in black During the period described in Fig10 therewere night-time low tides shortly after midnight on 14 and15 September 2006 when respectively both the LP-DOASand BBCRDS observed I2 substantially above their detec-tion limits and the model predicted peak I2 concentrations of50ndash100 pptv
Three modelled I2 datasets are presented in Fig10 (andother following data figures) covering the range of I2 re-cycling assumptions discussed in the previous section I2
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
11832 R J Leigh et al Bridging spatial scales in measurements of I2
Fig 11 Sources of emissions for 13 to 15 September 2006 Thetop panel shows total regional emissions data in the middle panelindicates the source of I2 emissions modelled at the measurementsite with the bottom panel showing the source of I2 emissions mod-elled along the LP-DOAS line of sight Modelled emissions in themiddle and bottom panels assume 95 recycling of photolysed I2during the daytime The emissions are coloured by seaweed speciesusing the convention of Fig7
concentrations modelled at the measurement site for ourusual assumption of 95 recycling of photolysed I2 are in-dicated by the orange line The lower boundary of the greyregion defines modelled I2 concentrations for theR = 090recycling assumption in line with results from theMahajanet al (2009) chemical recycling scheme The upper bound-ary of the grey region is defined by aR = 098 recyclingscheme The purpose of the grey region is to indicate rea-sonable boundaries of uncertainty in the model output in-troduced by recycling schemes which allow between 90 and98 of photodisocciated I2 to be reformed through IONO2chemistry During the night there is no photolysis and noI2 recycling via IONO2 chemistry and thus settingR to anyvalue 000ndash100 yields an identical result at night the greyregion collapses to the orange line alone Similar consider-ations apply to the blue line modelling the LP-DOAS mea-surements (forR = 095) and the accompanying grey region(090ltRlt098)
The dominant contributions fromL digitata andL hyperboreato the regional I2 emissions are highlighted inthe top panel of Fig11 The asymmetry of emission profilesthroughout the low tide cycle is also evident resulting fromthe decay in seaweed plantsrsquo I2 emission rates with timesince their first exposure to the atmosphere In contrast
Fig 12 Modelled and measured (BBCRDS) data from 16 Septem-ber 2006 with format as per Fig10
to the regional results in the top panel the contributionof each seaweed speciesrsquo emissions to the I2 measured bythe BBCRDS or LP-DOAS technique is highly dependentupon wind direction and speed (middle and lower panelsof Fig 11) For example the lack of a contribution fromL hyperboreato the BBCRDS modelled data is a commonfeature throughout the dataset resulting from the absenceof L hyperboreaseaweed beds in the shallower waters nearto the measurement site while the extensive and stronglyemitting L hyperboreabeds around the Ile de Batz and onthe coast west of Roscoff lie too far away to be included inthe footprint of air advected to the SBR site The LP-DOASdoes however have sensitivity toL hyperboreaemissionsfrom an area of this species growing to the south-west of theIle de Batz
Measured night-time concentrations of I2 around the pre-dawn low tide on 14 September (up to 50 pptv) are someof the highest recorded by the LP-DOAS instrument for thewhole campaign and are shown in Fig10 to be repro-duced well by the modelling results Agreement betweenthe BBCRDS observations and the model from 14 to 15September is also acceptable with the model reproducingthe main form of the BBCRDS measurements through to theearly hours of 15 Semptember (albeit the sharp I2 peak ofapprox 100 pptv due toL digitata emissions predicted bythe model towards the end of the time series is not present
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
R J Leigh et al Bridging spatial scales in measurements of I2 11833
Fig 13 Sources of emissions for 16 September 2006 with formatas per Fig11
in the BBCRDS measurement) The close correlation of theNO2 measurements by the NOxy and BBCRDS systems sug-gests good operation of both instruments during this periodand thus any measuredmodelled discrepancy for I2 is morelikely due to deficiencies of the model
Night-time measurements and model results are also il-lustrated in Figs12 and 13 for 16 September when re-gional emissions from predominantlyL digitata are pre-dicted around the evening-time low tide I2 concentrationsup to 25 pptv (significantly above the BBCRDS instrumentrsquoslimit of detection) were indeed seen around the low tide withthe measured I2 concentrations decreasing to zero at mid-night as the tide rises The model I2 concentrations peak atapproximately 20 pptv around low tide and then decay awayto zero in broad agreement with the measurement Howeverthe modelled I2 is more highly structured than the measure-ment indicating a high spatial dependence (ie wind direc-tion) of the I2 emissions reaching the measurement site (seealso middle panel of Fig13) Throughout the 16 Septemberdataset the BBCRDS NO2 measurements are again in excel-lent agreement with the NOxy chemiluminesence instrumentproviding confidence that the BBCRDS observational data isvalid The model also predicts sustained I2 emissions fromL ochroleucain the LP-DOAS line of sight (bottom panel ofFig13and middle panel of Fig12) though unfortunately noLP-DOAS measurements of I2 were available on this night
Measurements and modelled I2 concentrations from earlyin the campaign on 5 September are shown in Figs14and15 Although there is no NO2 data for this period (tocomment on the possible extent of I2 recycling) and no me-teorological data for the first few hours with which to pro-
Fig 14 Modelled and measured (LP-DOAS) data from 5 Septem-ber 2006 with format as per Fig10
Fig 15 Sources of emissions for 5 September 2006 with format asper Fig11
duce modelled I2 before 5 am this dataset illustrates a num-ber of interesting features Pre-dawn concentrations of I2 areboth predicted and seen along the LP-DOAS line of sightUp to 40 pptv of I2 is detected by LP-DOAS around the earlymorning low tide when the model also predicts up to 20 pptvof I2 before dawn and post-dawn for the the most extensiveR = 098 recycling regime The modelled and measured I2then both decrease to essentially zero as the tide rises from itsminimum in mid-morning through into the afternoon Afterdusk as the tide again recedes both modelled and measuredconcentrations increase significantly It should be noted thatemissions fromL hyperboreaare responsible for the largest
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
11834 R J Leigh et al Bridging spatial scales in measurements of I2
modelled concentrations at the end of this day (see Fig15)These are emitted into the LP-DOAS light path as it passesclose to theL hyperboreaseaweed bed south west of the Ilede Batz
62 Comparisons during the day
I2 photolyses rapidly during the day with I2 photolysis ratesreaching 025 sminus1 at solar noon (Fig4) corresponding toa lifetime of only 4 s for I2 Transport times from all butthe very closest seaweed beds are equivalent to many pho-tolytic lifetimes and thus photolytic destruction rapidly re-duces I2 concentrations modelled after dawn to negligibleamounts Indeed without a method for reforming photolysedI2 the model predicts that no I2 should be detectable eitherat the site or by the LP-DOAS instrument for the vast ma-jority of daylight hours (note for example the differencesbetween the modelled daytime I2 concentrations shown inpanels 2 and 3 and panels 4 and 5 of Fig 9) Howeverboth LP-DOAS and BBCRDS techniques clearly do mea-sure appreciable concentrations of I2 around most of the day-time low tides when the instruments were making I2 observa-tions suggesting a significant mechanism to reform I2 mustbe present The modelling of recycling schemes fromR = 90to 98 presents an opportunity to assess the likely extent of re-cycling required for the model to reproduce the observationsand to relate the extent of recycling to the NO2 concentra-tions co-measured by the BBCRDS and NOxy instruments
Modelled concentrations and observational data aroundthe daytime low tide on 25 September are shown in Fig16Rather low concentrations of I2 are predicted on this daywith concentrations reported by the BBCRDS instrumentconsistently below its detection limit for I2 including atthe tidal minimum Some of the lowest NO2 concentra-tions of the whole campaign were also measured by theBBCRDS and NOxy instruments on this day being approx-imately 05 pptv for the duration of the BBCRDS observa-tions (bottom panel) This is significantly below the 2 pptv ofNO2 considered by theMahajan et al(2009) chemical modelof I2 recycling via IONO2 which suggested the recycling pa-rameter was aroundR = 090 For the particularly clean con-ditions (for this location) of 25 September with NO2 sig-nificantly below 2 pptv recycling is likely to be ineffectiveat offsetting the rapid photolytic losses of I2 Therefore themost reasonable modelled I2 concentrations are likely to liebelow the grey region (ie below even theR = 090 assump-tion) commensurate with the low I2 amounts reported bythe BBCRDS instrument The exception is the spike of 25ndash75 pptv I2 (depending on whereR is assumed in the range090 to 098) due to a sharp rise in modelledL digitataemis-sions received at the site caused by a short-lived shift inwind direction (see Fig17) but which is not evident in theBBCRDS measurements themselves
Fig 16 Modelled and measured (BBCRDS) data from 25 Septem-ber 2006 with format as per Figure10
Fig 17 Sources of emissions for 25 September 2006 with formatas per Fig11
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
R J Leigh et al Bridging spatial scales in measurements of I2 11835
Fig 18Modelled and measured (LP-DOAS) data from 19 Septem-ber 2006 with format as per Fig10
Daytime data from 14 September included in Fig10and introduced above indicate broad agreement betweenBBCRDS measurements and modelled concentrations at sitewithin the boundaries established by theR = 090 to 098 re-cycling schemes It should be noted here that although NO2concentrations at site experienced a local minimum of under1 ppbv around the time of the highest BBCRDS I2 concen-trations the usual NO2 concentrations on this day are sub-stantially higher than on clean days such as 25 September(Fig 16) and significant recycling is still needed to bringthe model into agreement with the BBCRDS measurementsThe modelled I2 concentrations during daylight hours on theprevious day 13 September are generally smaller than theLP-DOAS measurements (see middle panel of Fig10) al-though there are spikes in the modelled I2 around sunset of asimilar size to the observations Indeed there is a particularlylarge amount of variability in the modelled and measured I2concentrations measured NO2 and emission sources on 13September most likely due to highly variable wind condi-tions on this day
For completeness LP-DOAS data from 19 September areshown in Fig18 There is generally good agreement be-tween the data and model during the first phase of this timeperiod ie a very low level of I2 is observed from midnightuntil the tide ebbs around dawn when 10 pptv of I2 is ob-served and modelled There is then poor agreement for themajority of the day despite high data density and relatively
low error estimates on the LP-DOAS data the LP-DOASsystem detects I2 well above its detection limit throughoutdaylight hours which is not reproduced in the model
63 Discussion of differences
This modelling activity has demonstrated a number of strongcorrelations between modelled and measured I2 concentra-tions for both measurement geometries and provided infor-mation on the likely I2 sources (seaweed speciation and ge-ographical location) However significant discrepancies re-main most notably those on 19 September for the longestLP-DOAS data sequence indicating inaccuracies in modelinputs andor more fundamental limitations of the modellingapproach applied in this work
Potentially inaccurate input data include the spatial dis-tribution of seaweed species including small patches of sea-weed not accurately represented by or missing from the origi-nal seaweed maps (Braud 1974 Bajjouk et al 1996) mixedseaweed beds containing more than one species and anyloose seaweed that had washed up on the shore near to mea-surement site Although previous studies were used to pro-duce an updated map for the present modelling study anaerial survey coupled with further surface studies would re-duce uncertainties in this area Furthermore there are noI2 emission data available forL ochroleucafrom previouslaboratory studies Here this species was assumed to emitat a rate intermediate betweenL digitataL hyperboreaandSaccharina lattisima but if this assumption is in error itcould have a significant impact on modelled I2 amounts par-ticularly for the LP-DOAS measurement geometry whereL ochroleucais the dominant emission source for certainmeteorological conditions (see Fig13 for example) Alsothe spatial variability of wind fields across the eulittoral zoneis not considered in detail in this study Although two in-dependent measurements of wind speed and direction weretaken at the BBCRDS measurement site and showed strongagreement the local scale topography is likely to have causeda some variation in wind vectors within the eulittoral zoneIn particular variability in wind vectors increase uncertain-ties in the contributions made by theLaminariabeds aroundthe Ile de Batz For example emissions from these seaweedsare responsible for the large spikes in modelled I2 concen-trations above 100 pptv shown in Fig9 and which are gen-erally not replicated in the measurements (ie the modelledI2 footprints are likely to be too directional even withplusmn 5degree uncertainty already assumed in the model)
Fundamental limitations in our the relatively simplisticmodelling include the parameterisation of seaweed mass perunit sea surface area and characterisation of the surfaceroughness and slope in the footprint modelling Drawingconclusions about the extent of daytime I2 recycling and itsrelationship with NO2 concentrations also rely on an under-standing of the concentration of NO2 above the eulittoralzone and along the full transport path of I2 A comprehensive
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
11836 R J Leigh et al Bridging spatial scales in measurements of I2
modelling treatment for this factor demands a more advancedchemical scheme and a robust understanding of iodine chem-istry in the semi-polluted environment (Mahajan et al 2009)As with the wind vectors NO2 concentrations measuredat site are not fully representative of NO2 concentrationsthroughout the modelled region of interest with its many var-ied NOx sources
Of the potential sources of error discussed here we sug-gest the major contributors to modelmeasurement differ-ences are inaccuracies in mapping the spatial distributionsof seaweed habitats (particularly for seaweeds close to mea-surement locations) uncertainties inL ochroleucaemissionrates and the variability in wind vectors along the transitpath from emission source to detection Given the closeagreement between NO2 measurements from the NOxy andBBCRDS instruments and the established pedigree of LP-DOAS as an atmospheric measurement technique coupledwith high data density and low error bars calculated from theDOAS fits inaccuracies in the measurements are not consid-ered to be a significant reason for any discrepancy with themodelled data
7 Conclusions
A dynamical model was produced to examine the sensitivityof in situ and line of sight measurements based at the Sta-tion Biologique de Roscoff to regional emissions of molec-ular iodine during the RHaMBLe campaign in September2006 Modelled concentrations of I2 were compared to mea-sured concentrations from a BBCRDS instrument located onthe shore and a LP-DOAS instrument with an absorptionpath extending over the eulittoral zone Although havingsimplifying assumptions this model nevertheless provides ameans to assess likely impacts on measured I2 concentrationsfrom time-dependent variations in emissions from the differ-ent seaweed species growing in the area and the spatial loca-tion of seaweed beds relative to the measurement locationsFurthermore the boundaries of feasible I2 recycling schemessuggested by our measurements and by recent chemical mod-elling studies were explored
Using previous laboratory measurements of species-dependent I2 emission rates concentrations of I2 above themost strongly emittingL hyperboreaand L digitata bedswere predicted to be 15 ppbv immediately after being firstexposed to the atmosphere by a retreating tide in line with re-cent laboratory measurements (Ball et al 2010) Although afew spikes above 100 pptv were predicted concentrations atthe BBCRDS measurement site and along the LP-DOAS lineof sight were generally modelled to be below 50 pptv andshow some quantitative agreement with measured datasetsThis demonstrates the use of concentration footprints in thiscontext to explore the substantial dispersion and dilutionalong a transit path linking the high concentrations directlyabove emission sources to the much lower concentrationsmeasured downwind
Modelled concentrations during the day were demon-strated to be highly sensitive to recycling schemes allow-ing the reformation of photodissociated I2 Without any re-cycling mechanism concentrations of I2 modelled for bothmeasurement geometries would usually have been negligi-ble in marked contrast to the BBCRDS and LP-DOAS obser-vations themselves which both report clear daytime I2 signalsabove their detection limits (approx 10 pptv) on a number ofoccasions during the campaignMahajan et al(2009) mod-elled I2 concentrations measured by the LP-DOAS instru-ment during RHaMBLe and also concluded that it was nec-essary to invoke substantial recycling of I2 Further chem-ical modelling was performed to support the present studyre-running theMahajan et al(2009) model for different I2photolysis rates with and without NOx present It was foundthat the decrease in I2 concentrations downstream of an emis-sion source (using a baseline NO2 concentration of 2 ppbv)was well matched to the results of re-running theMahajanet al(2009) model with the NOx chemistry ldquoturned offrdquo butwith an I2 photolysis rate reduced to only 10 of its usualvalue Hence a very simplistic recycling scheme was intro-duced into the present model (which has no NOx chemistry)to mimic the effects of I2 recycling in a semi-polluted NOxenvironment by reducing the effective I2 photolysis frequen-cies starting with a recycling parameter ofR = 090 (ie theI2 photolysis rate is reduced to 10 of its measured value assuggested by the extraMahajan et al(2009) modelling)
Employing a recycling rate ofR = 090 produced modelledI2 concentrations that were still generally below 5 pptv dur-ing daylight hours Except for clean condition encounteredon one day (25 September [NO2] less than 05 ppbv) whengood quality BBCRDS data consistently showed I2 concen-trations below 10 pptv recycling rates ofR = 095 to 098needed to be included in the model to bring the modelledI2 into agreement with the observed I2 amounts It shouldbe noted that more extensive I2 recycling (ieR gt 090) isexpected when NO2 concentrations are above 2 ppbv (whichwas often the case during RHaMBLe) as a greater fraction ofiodine will be converted into the temporary IONO2 reservoirto act as a potential downstream source of I2 The greater re-cycling efficiencies required in the present model do not nec-essarily mean I2 recycling chemistry is even more extensivethan proposed in the originalMahajan et al(2009) studyit may just be a consequence of the simplistic scheme ourmodel has used to parameterise I2 recycling
However the semi-polluted NOx regime at Roscoff is asnoted byMahajan et al(2009) rather different from that forprevious I2 measurements at cleaner background sites mostnotably at the Mace Head Atmospheric Research Station(Galway Ireland) where typical NO2 concentrations (Heardet al 2006) are below the minimum NO2 levels observedduring RHaMBLe and thus little recycling of I2 via IONO2chemistry is expected
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
R J Leigh et al Bridging spatial scales in measurements of I2 11837
Analysis such as that produced in this paper is recom-mended to aid the interpretation of results from in situ mea-surements performed in complex environments containingspatially inhomogeneous and temporally varying emissionfields Ideally the framework presented in this paper wouldin future be coupled with a suitable chemistry scheme tomodel I2 observations with a full treatment of the NOx de-pendent recycling chemistry An enhanced scheme couldalso then be used to model other observables in iodinersquos ma-rine boundary layer chemistry most importantly the IO rad-ical and iodine-induced particle nucleation (von Glasow andCrutzen 2007 McFiggans et al 2010) The formation ofIO from I2 (via I2 photolysis and reaction of the resultingI atoms with O3) happens rapidly on the timescale of 10 sduring the day whereas nucleation of particles from IOrarrOIOrarr viable nucleirarr detectible particles inevitably oc-curs over longer timescales Thus these processes are likelyto be affected differently by variable meteorology and spa-tially inhomogeneous emission fields even for co-located I2IO and particle measurements
Total regional emissions for the 100 km2 zone aroundRoscoff have been modelled to be up to 17times 1019 moleculesper second during the lowest tides The dominant con-tribution to regional emissions is predicted to derive fromL hyperboreaand to a lesser extentL digitata from a scal-ing of laboratory-based studies of I2 emission rates from thedifferent seaweed species and speciated maps of their habi-tats around Roscoff Whereas the major contributor to thenet regional emissions isL hyperborea this modelling hasshown that the majority of the BBCRDS and LP-DOAS I2signals derive fromL digitata andL ochroleuca Althoughthe L hyperboreaseaweed beds are too far away from theSBR and LP-DOAS light path to (usually) be included inthe footprints of air advected to the measurement locationsL hyperboreaandL digitataemissions are almost certainlystill the dominant source of newly nucleated aerosol parti-cles since (i) these seaweed species are the strongest emitters(Ball et al 2010) and (ii) particle nucleation is a non-linearprocess occurring in ldquohot spotsrdquo of locally elevated concen-trations of iodine oxides (Burkholder et al 2004) Thus spa-tial inhomogeneity in the emission field is likely to be evenmore influential on the kinetics of particle nucleation than forthe I2 observations discussed in this work
This study has illustrated the challenge of combining ob-servational data from point andor line sensors with emissionmaps to produce a metric representative of the net regionalemissions without extensive spatially-resolved dynamicaland chemical modelling These findings have been demon-strated for the marine coastal environment in this study butare equally applicable in any other scenario in which emis-sions are spatially inhomogeneous and temporally variable(eg multiple emission sources in the urban environment)The value of long-path DOAS techniques that measure ab-sorber amounts over an extended air mass already provide theability to survey multiple emission sources within their line
of sight In the future such advantages could be extended bydeploying open-path DOAS systems that use scattered sun-light and can scan the direction of their field of view over theentire measurement region to build a map of absorber con-centrations provided that appropriate detection limits and theability to isolate local emissions could be assured
AcknowledgementsThe authors would like to thank the staff at theStation Biologique de Roscoff for their significant assistance duringthe RHaMBLe project and the Natural Environment ResearchCouncil for funding the RHaMBLe campaign Deployment ofthe BBCRDS instrument to the RHaMBLe campaign was madepossible through a grant from the Natural Environment ResearchCouncil NED00652X1
Edited by E Pelinovsky
References
Arzel P Les laminaires sur les cotes bretonnesevolution delrsquoexploitation et de la flottille de pecheetat actuel et perspec-tives Edition de lrsquoIfremer p 139 1998
Bajjouk T Guillaumont B and Populus J Application of air-borne imaging spectrometry system data to intertidal seaweedclassification and mapping Hydrobiologia 327 463ndash471 1996
Ball S M and Jones R Broad-band cavity ring-down spec-troscopy Chem Rev 103 5239ndash5262 2003
Ball S M and Jones R Broadband cavity ring-down spec-troscopy in ldquoCavity ring-down spectroscopy Techniques andapplicationsrdquo edited by Berden G and Engeln R BlackwellPublishing Ltd 2009
Ball S M Hollingsworth A M Humbles J Leblanc C PotinP and McFiggans G Spectroscopic studies of molecular iodineemitted into the gas phase by seaweed Atmos Chem Phys 106237ndash6254 doi105194acp-10-6237-2010 2010
Bitter M Ball S Povey I and Jones R A broadbandcavity ringdown spectrometer for in-situ measurements of at-mospheric trace gases Atmos Chem Phys 5 3491ndash3532doi105194acp-8-3491-2005 2005
Braud J-P Etude de quelques parametres ecologiques bi-ologiques et biochimiques chez une pheophycee des cotes bre-tonnes Laminaria ochroleuca Revue des Travaux de lrsquoInstitut desPeches Maritimes (ISTPM) 38 1974
Burkholder J B Curtius J Ravishankara A R and Love-joy E R Laboratory studies of the homogeneous nucleationof iodine oxides Atmos Chem Phys 4 19ndash34 doi105194acp-4-19-2004 2004
Dixneuf S Ruth A A Vaughan S Varma R M and Or-phal J The time dependence of molecular iodine emissionfrom Laminaria digitata Atmos Chem Phys 9 823ndash829 doi105194acp-9-823-2009 2009
Edwards G D and Monks P Performance of a single monochro-mator diode array spectroradiometer for the determination of ac-tinic flux and atmospheric photolysis frequencies J GeophysRes 108 8546 2003
Gevaert F Janquin M-A and Davoult D Biometrics in Lami-naria digitata a useful tool to assess biomass carbon and nitro-gen contents J Sea Res 60 215ndash219 2008
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
11838 R J Leigh et al Bridging spatial scales in measurements of I2
Gollety C Migne A and D D Benthic metabolism on a shel-tered rocky shore role of the canopy in the carbon budget JPhycol 44 1146ndash1153 2008
Heard D E Read K A Methven J Al-Haider S Bloss W JJohnson G P Pilling M J Seakins P W Smith S C Som-mariva R Stanton J C Still T J Ingham T Brooks BDe Leeuw G Jackson A V McQuaid J B Morgan RSmith M H Carpenter L J Carslaw N Hamilton J Hop-kins J R Lee J D Lewis A C Purvis R M WevillD J Brough N Green T Mills G Penkett S A PlaneJ M C Saiz-Lopez A Worton D Monks P S FlemingZ Rickard A R Alfarra M R Allan J D Bower KCoe H Cubison M Flynn M McFiggans G GallagherM Norton E G OrsquoDowd C D Shillito J Topping DVaughan G Williams P Bitter M Ball S M Jones R LPovey I M OrsquoDoherty S Simmonds P G Allen A Kinner-sley R P Beddows D C S DallrsquoOsto M Harrison R MDonovan R J Heal M R Jennings S G Noone C andSpain G The North Atlantic Marine Boundary Layer Exper-iment(NAMBLEX) Overview of the campaign held at MaceHead Ireland in summer 2002 Atmospheric Chemistry andPhysics 6 2241ndash2272 doi105194acp-6-2241-2006 2006
Luning K Seaweeds Their environment biogeography and eco-physiology Wiley 1990
Mahajan A Oetjen H Saiz-Lopez A Lee J D McFiggansG B and Plane J M C Reactive iodine species in a semi-polluted environment Geophys Res Lett 36 L16803 doi1010292009GL038018 2009
Martin F Bacis R Churassy S and Verges J Laser-induced-fluorescence Fourier transform spectrometry of the X16+g state
of I2 Extensive analysis of the B35+u rarrX16+g fluorescence
spectrum of127I2 J Molec Spectrosc 116 71 1986McFiggans G Coe H Burgess R Allan J Cubison M Al-
farra M R Saunders R Saiz-Lopez A Plane J M CWevill D Carpenter L Rickard A R and Monks P SDirect evidence for coastal iodine particles from Laminariamacroalgae ndash linkage to emissions of molecular iodine AtmosChem Phys 4 701ndash713 doi105194acp-4-701-2004 2004
McFiggans G Bale C S E Ball S M Beames J M Bloss WJ Carpenter L J Dorsey J Dunk R Flynn M J FurneauxK L Gallagher M W Heard D E Hollingsworth A MHornsby K Ingham T Jones C E Jones R L KramerL J Langridge J M Leblanc C LeCrane J-P Lee J DLeigh R J Longley I Mahajan A S Monks P S OetjenH Orr-Ewing A J Plane J M C Potin P Shillings A JL Thomas F von Glasow R Wada R Whalley L K andWhitehead J D Iodine-mediated coastal particle formationan overview of the Reactive Halogens in the Marine BoundaryLayer (RHaMBLe) Roscoff coastal study Atmos Chem Phys10 2975ndash2999 doi105194acp-10-2975-2010 2010
Plane J M C and Saiz-Lopez A Analytical techniques for atmo-spheric measurement Blackwell 2006
Platt U Modern methods for the measurement of atmospherictrace gases Phys Chem Chem Phys 1 5409ndash5415 1999
Saiz-Lopez A and Plane J M C Novel iodine chemistry in themarine boundary layer Geophys Res Lett 31 L04 112 doi1010292003GL019215 2004
Saiz-Lopez A Saunders R Joseph D M Ashworth S H andPlane J M C Absolute absorption cross-section and photolysisrate of I2 Atmos Chem Phys 4 1443ndash1450 doi105194acp-4-1443-2004 2004
Saiz-Lopez A Plane J M C McFiggans G Williams P IBall S M Bitter M Jones R L Hongwei C and HoffmannT Modelling molecular iodine emissions in a coastal marineenvironment the link to new particle formation Atmos ChemPhys 6 883ndash895 doi105194acp-6-883-2006 2006
Schmid H P Source areas for scalars and scalar fluxes BoundaryLayer Meteorology 67 293ndash318 1994
Shillings A Atmospheric applications of broadband cavity ring-down spectroscopy PhD Thesis University of Cambridge2009
Vandaele A Hermans C Simon P Van Roozendael M Guil-mot J Carleer M and Colin R Fourier transform measure-ment of NO2 absorption cross-sections in the visible range atroom temperature J Atm Chem 25 289ndash305 1996
von Glasow R and Crutzen P Tropospheric halogen chem-istry in Treatise on Geochemistry edited by Hol-land H D and Turekian K K Pergamon Oxford 1ndash67doidoi101016B0-08-043751-604141-4 httpwwwsciencedirectcomsciencearticleB782S-4CJV6M2-15235ca9af61527f9815d05a526b4673865 2007
Western C PGOPHER a program for simulating rotational struc-ture available University of Bristolhttppgopherchmbrisacuk Access September 2009
Zilitinkevich S S On the computation of the basic parameters ofthe interaction between the atmosphere and the ocean Tellus21 17ndash24 1969
