JOURNAL OF GEOPHYSICAL RESEARCH, VOL. ???, NO. , PAGES 1–28,
Stratospheric Variability and Tropospheric1
Ozone2
Juno Hsu and Michael J. Prather
Juno Hsu, Earth System Science, University of California, Irvine. 92697. Email: [email protected]
1Earth System Science, University of
California, Irvine, California, USA
D R A F T September 3, 2008, 12:07pm D R A F T
2 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
Abstract.
Changes in the stratosphere-troposphere exchange (STE) of ozone are ex-
pected to alter the tropospheric ozone abundance, both over the last decades
of stratospheric ozone depletion and the coming century of climate change.
Combining an updated linearized stratospheric ozone chemistry (Linoz v2)
with parameterized PSCs and a five-year sequence of EC meteorology we ex-
amine the variations in STE O3 flux and how they perturb tropospheric O3.
Our best estimate for the current STE ozone flux is 290 Tg/yr in the NH
and 225 Tg/yr in the SH with interannual variability over years 2001-2005
of ± 25 Tg/year and ± 30 Tg/year, respectively. The STE flux alone drives
a large seasonal change in the tropospheric ozone burden with NH mid-latitude
peak-to-peak changes of about 8 DU, mimicking summertime photochem-
ical production but with half the amplitude. The model matches the quasi-
biennial oscillation (QBO) in column ozone, and the STE shows negative anoma-
lies over the mid-latitudes during the easterly phases of the QBO and vice
versa. The QBO-induced circulations over mid-latitudes during the easterly
phase create conditions that reduce STE. The tropospheric burden of this
O3 of stratospheric origin is indeed linear with STE. When the observed col-
umn ozone depletion from 1979 to 2004 is modeled with Linoz v2, we pre-
dict STE reductions of at most 10 % in NH, corresponding to about 1 ppb
decrease in hemispheric tropospheric O3.
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 3
1. Introduction
Scientific e!orts to understand the trends and variations in ozone observed over the past3
few decades has demonstrated the role of both photochemical and meteorological factors4
in driving stratospheric ozone change (e.g., Randel and Wu, 2007; Stolarski et al., 2006,5
Salawitch et al., 2005). It has been proposed that these stratospheric changes have altered6
the tropospheric ozone burden over the past few decades (Fusco and Logan, 2003) and7
will continue to a!ect it in the future (Sudo et al., 2003). This paper presents a series of8
highly constrained modeling experiments that capture the observed trends and variations9
in stratospheric ozone and diagnose the corresponding changes in the stratosphere-to-10
troposphere flux of ozone. We are thus able to better understand the seasonal, interannual,11
and decadal trends in tropospheric ozone and the oxidative capacity of the atmosphere12
that are driven by the stratosphere.13
The coupling of stratospheric and tropospheric ozone with chemistry models or with14
chemistry-climate models is occurring across the community (Eyring et al., 2005). These15
full models include a nearly complete set of chemical species and reactions that a!ect16
ozone, but are costly to run, and are often di"cult to diagnose as to the causative fac-17
tors of variability. We approach the problem with a simplified chemical model that is18
focused on simulating the stratosphere-to-troposphere exchange (STE) of ozone: a lin-19
earized ozone chemistry (Linoz version 1: McLinden et al., 2000) combined with unique20
transport diagnostics that quantify the STE flux as a function of time and place (Hsu21
et al., 2005). The Linoz model is revised (Section 2) to use an updated climatology22
for the background stratospheric composition and current photochemical data (IUPAC,23
2004; Sander et al., 2006). Stratospheric ozone simulated with the new Linoz version 224
D R A F T September 3, 2008, 12:07pm D R A F T
4 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
driven by Oslo ECMWF-IFS meteorology data is tested against observed ozone climatol-25
ogy (Section 3). The seasonal and interannual relationship between stratospheric ozone,26
STE flux, and tropospheric ozone is derived for a continuous sequence of meteorological27
fields from January 2000 to December 2005 (Section 4). The extent of the halogen-driven28
ozone STE decrease since 1979 is derived (Section 5), and we discuss the overall role of29
the stratosphere in driving tropospheric ozone change (Section 6).30
We find that stratosphere alone produces a peak-to-peak seasonal variation in tropo-31
spheric column ozone of about 8 DU at northern mid-latitudes that mimics tropospheric32
photo-chemistry. For longer time scales, the Quasi-Biennial Oscillation (QBO) signature33
in total column ozone is roughly matched to the observed and the QBO signals in ozone34
STE flux have maximum amplitudes in midlatitudes that are opposite in phase to its35
midlatitude QBO signal in total ozone. The observed, post-1979 ozone depletion for the36
NH can be best simulated in our chemistry-transport model (CTM) with a 4K higher37
threshold activation temperature than the typical 195K threshold for PSC formation.38
Enhanced background bromine levels are found to have negligible e!ect on ozone deple-39
tion but our PSC chemistry is parameterized for fixed bromine and so only our gas-phase40
chemistry responds to enhanced Bry. The maximum simulated decrease in the STE flux41
for post-1979 ozone depletion is about 10% in the northern hemisphere (NH) and 22% in42
the southern hemisphere (SH). Furthermore, the latitude-season pattern of STE decrease43
due to ozone depletion is distinctly di!erent from the change in total column ozone.44
2. A linearized stratospheric Ozone Chemistry – Linoz version 2
The Linoz tables are derived using the photochemistry box model of Prather (1992) with45
background atmospheric composition specified as a monthly, latitude-height climatology46
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 5
from observations. The local net photochemical production (P-L) is represented by a47
first-order Taylor Series expansion using only three independent variables: local ozone48
mixing ratio (f), temperature (T), and overhead ozone column (c),49
df
dt= (P ! L)o +
!(P ! L)
!f
!!!!o
(f ! f o) +!(P ! L)
!T
!!!!o
(T ! T o) +!(P ! L)
!c
!!!!o
(c ! co). (1)
The photochemical tendency for the climatology is denoted by (P-L)o ,and the clima-50
tology values for the independent variables are denoted by fo , co , and To , respectively.51
Including these four climatology values and the three partial derivatives, Linoz is defined52
by seven tables. Each table is specified by 216 atmospheric profiles: 12 months by 1853
latitudes (85S to 85N). For each profile, the net production for the climatology, (P-L)o ,54
and the three derivatives are evaluated at every 2 km in pressure altitude from z* = 1055
to 58 km (z* = 16 km log10 (1000/p)). These tables are automatically remapped onto56
any CTM grid with the mean vertical properties for each CTM level calculated as the57
mass-weighted average of the interpolated Linoz profiles.58
We adopt the ozone climatology compiled by McPeters et al. (2007), which has im-59
proved profiles over the tropics and the SH as compared with Linoz v1. The temperature60
climatology is unchanged (Nagatani and Rosenfield, 1993). The remaining chemical com-61
position is specified as a climatology scaled to the tropospheric abundance of the long-lived62
source gases (i.e., N2O, CH4, and the halocarbons) so that it can be changed to reflect63
a changing atmosphere. This includes climatologies for three chemical families (NOy =64
NO + NO2 + HNO3 + ...; Cly = ClO + HOCl + ClONO2 + HCl + ...; Bry = Br +65
BrO + BrONO2 + ...). We use N2O as the primary measure of stratospheric composition66
and tracer-tracer relations to define the other trace gases. A monthly, latitude-by-height67
D R A F T September 3, 2008, 12:07pm D R A F T
6 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
N2O climatology above 22 km is based on CLAES satellite measurements (Oct 1991- May68
1993, Randel et al, 1994) and below 22 km is constructed from the compact correlation69
with O3 from the NASA ER-2 in-situ measurements in the lower stratosphere (Strahan70
et al., 1999). Minor smoothing is applied to the transition region. The CH4 and NOy71
distributions are obtained using the polynomial fit with respect to N2O from ATMOS mea-72
surements (Michaelsen et al., 1998a; 1998b). The Cly climatology assumes conservation73
of halogens, thus increasing in the stratosphere as the organic source gases (e.g., CFCs,74
CH3CCl3, CCl4) are photochemically destroyed (Woodbridge et al., 1995). The Bry cli-75
matology likewise assumes increasing Bry as the tropospheric bromine source gases (e.g.,76
CH3Br, CF2BrCl, CF3Br) are destroyed (Wamsley et al., 1998). For Bry, we consider a77
sensitivity case where the tropopause value is increased to 6 ppt to include the relatively78
large amounts of inorganic bromine (Bry) that may cross the tropopause (Salawitch et79
al., 2005). Both families are keyed to the N2O distribution. Water vapor adopts a lower80
boundary fixed at 3.65 ppm (Nassar et al., 2005) and conserves of total hydrogen (H2O +81
2xCH4). The tracer-tracer correlations are applied with N2O scaled to the year of their82
observed correlations. Normalized distribution patterns for N2O, CH4, H2O, NOy and83
Bry in January are shown in Fig. 1. Compared to v1, these Linoz v2 climatologies for84
background stratospheric composition more accurately match observations.85
Ozone and temperature climatologies, to first order, determine stratospheric photolysis86
rates. We adopt a surface reflectivity of 0.3 as an average cloud cover. The photochem-87
ical box model is initialized with an approximate balance of species within each of the88
chemical families and integrated for 30 days to reach an approximate, diurnally repeat-89
ing steady state, whereby the initialization of species within the families is, for the most90
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 7
part, forgotten. During this integration, the abundance of ozone and long-lived gases91
are fixed, and the chemical families are conserved. The net ozone production and three92
partial derivatives are evaluated numerically by perturbing the local ozone by +5%, the93
column ozone by +5%, and the temperature by +4 K. Regarding the nonlinearity in the94
derivatives, see Fig. 3 of McLinden et al (2000).95
For Linoz version 2, the photochemistry has been updated from the 1997 -vintage version96
1 to current rate coe"cients (Sander et al., 2006) and cross sections (IUPAC, 2004). New97
solar fluxes are taken from the average solar irradiance reference spectra derived by the98
SUSIM team for two di!erent levels of of solar activity (Thuillier et al., 2004). Compared99
to Linoz v1, these are 15-20 % larger at wavelengths 177-200 nm, 5-10% larger at 200-100
300 nm, but relatively unchanged long-ward of 300 nm. Other notable updates a!ecting101
photolysis rates include the quantum yield of O(1D) from O3 photolysis and the NO2 cross102
sections.103
As an example of how the stratospheric chemistry model has evolved since v1, we follow104
the chemistry updates using a standard ATMOS profile (May 31, 30N) from previous105
models and measurements studies (Prather and Remsberg, 1993). The height profiles of106
net ozone production (P-L) and its derivatives with respect to ozone, temperature, and107
column ozone are shown in Fig. 2. A sequence of six model calculations are shown with108
successive updates tracking the change in chemistry from Linoz v1 to v2. Values generated109
with the JPL-1997 kinetics rates and cross sections and with the old solar flux data used110
for generating Linoz v1 are shown for comparison (JPL97-S3). Updating the quantum111
yields and cross sections only (JPL97-S2) has no e!ect on the three derivatives and a112
barely noticeable e!ect on net production, i.e., a small increase near 40 km. Updating113
D R A F T September 3, 2008, 12:07pm D R A F T
8 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
the solar fluxes in addition (JPL97-S1) also has no e!ect on the derivatives but causes a114
large increase in net production throughout the stratosphere above 25 km. The update115
to JPL 2000 kinetics (JPL00-S1) causes a notable decrease in the temperature derivative116
between 34 and 48 km with an increase in net production from 34 to 44 km and a decrease117
below 34 km. Updating the kinetics to JPL 2002 (JPL02-S1) and JPL 2006 (JPL06-S1)118
has minor e!ects, with the latter causing a small decrease in net production about 30 km.119
The largest and most extensive changes in the chemical model occur in the net ozone120
production and not in the derivatives. The largest change in updating from JPL 1997121
to JPL 2000 is caused by the addition of a new branch for the reaction, OH + ClO "122
HCl +O2. This new pathway weakens the Cly-catalyzed ozone loss and thus results in123
an increase in net production peaking around 38 km. The other major change with JPL124
2000 kinetics was a stronger NOy-catalyzed ozone loss from the increased kinetic rate for125
the reaction, NO2 + O " NO + O2, and decreased kinetic rate for the reaction, NO2 +126
OH " HNO3. Changes to chemical reaction rates after JPL 2000 have relatively minor127
e!ects on the ozone chemistry (outside of PSC conditions).128
Linoz v1 considered only gas-phase photochemistry and did not include chlorine acti-129
vation by Polar Stratospheric Clouds (PSCs). Thus, in v1 there was no Antarctic ozone130
hole and no enhanced Arctic loss during cold winters. In v2, we incorporate the PSC131
parameterization scheme of Cariolle et al. (1990) when the temperature falls below 195132
K and the sun is above the horizon at stratospheric altitudes. The O3 loss scales as the133
squared stratospheric chlorine loading (normalized by the 1980 level threshold). In this134
formulation PSC activation invokes a rapid e-fold of O3 based on a photochemical model,135
but only when the temperature stays below the PSC threshold. It does not consider that136
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 9
the activated chlorine continues to destroy ozone for several days after encountering a137
PSC (Schoeberl et al., 1993). Recently, Cariolle and Teyssedre (2007) added a cold-tracer138
to account for this e!ect. Their new parameterization, which is not used here, will be139
more important in the Arctic where PSCs are not sustained throughout the winter. In140
view of this process and the evidence of Cly activation on ternary aerosols at warmer tem-141
peratures (Thornton et al., 2005), we test another version using a higher PSC-activation142
temperature of 199K.143
Linoz chemical tendencies are applied only in the stratosphere, defined here as to CTM144
grid points for which the O3 abundance is greater than 100 ppb. These simulations do145
not include realistic tropospheric ozone chemistry, but instead invoke a parameterized146
sink that restores O3 to 20 ppb in the lowest 600 m of the troposphere with an e-folding147
time-scale of 2 days (Hsu et al., 2005). The choice of 20 ppb was made to imitate a more148
realistic chemistry and produce reasonable tropospheric column O3. This tropospheric149
pseudo-chemistry is uniform, and thus variations in tropospheric O3 calculated here are150
driven entirely by the STE flux. When combining Linoz with a full tropospheric chemistry151
model, we simulate a separate Linoz tracer (O3s) and use it every time step in each grid152
box to determine if the tropospheric chemistry is invoked (e.g., O3s <100 ppb) or if the153
Linoz net chemical tendencies are used (>100 ppb).154
Using the normalized, monthly 2-D climatologies for stratospheric composition, we cal-155
culate five sets of Linoz v2 tables (see Table 1). Linoz-1979 uses the 1979 mean abundances156
from REF 1 of Eyring et al. (2005) and represents a stratosphere prior to significant157
ozone depletion. With Cly levels below the chlorine-loading threshold, PSC-induced loss158
is never invoked with Linoz-1979. Linoz-2004 uses year-2004 mean tropospheric abun-159
D R A F T September 3, 2008, 12:07pm D R A F T
10 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
dances (WMO, 2006, Table 1-2) and generates an Antarctic ozone hole. A second pair of160
Linoz tables, -1979Br and -2004Br, assume a 6ppt greater background of Bry throughout161
the stratosphere (see Salawitch et al., 2005). We also use the Linoz-2004Br tables with162
a warmer PSC threshold of 199K and denote this case as Linoz-2004BrT. Note that the163
Linoz tables assume only gas phase chemistry plus some sulfate reactions using a SAGE164
climatology for the aerosol surface area. Thus ClO levels in the lower stratosphere are165
always low and the enhanced Bry does not notably enhance ozone loss.166
3. Evaluating Column Ozone with Linoz
To assess the impact of the updated Linoz v2 on stratospheric ozone, we repeat the167
Linoz v1 simulations of Hsu et al. (2005) with Linoz-2004 and the Oslo/EC meteorology168
for year 1997 derived from the European Centre for Medium-Range Weather Forecasts169
(ECMWF) Integrated Forecasting System (IFS) Cycle 23r4. Linoz v1 is known to be170
biased low in column O3 in the tropics and high in high latitudes (see Fig. 1 of Wild171
et al., 2003). With Linoz-2004 this bias is mostly eliminated: tropical ozone columns172
increase by 5-20% for all but the northern winter, and outside of the tropics ozone is173
reduced by similar percentages for all months except December. In terms of STE, if we174
run Linoz-2004 tables but turn o! the PSC parameterization, the O3 flux increases by 9%175
from 516 Tg/yr to 563 Tg/yr, with greater increases in the SH. The spatial and temporal176
STE patterns remain roughly the same. Inclusion of the parameterized PSC chemistry177
with Linoz-2004 reduces the STE fluxes globally by 10%, again with greater response in178
the SH. The total shift in STE flux from v1 to v2 (Linoz-2004) is +3% in the NH and179
-7% in the SH. The changes in the photochemical data and inclusion of a parameterized180
PSC loss have corrected the prominent biases in Linoz v1.181
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 11
Stratospheric O3 columns calculated with Linoz-2004 with 1997 ECMWF IFS Cycle23r4182
meteorological data are compared with those from observations (McPeters et al., 1997;183
2007) in Fig. 3. We show results for both with and without PSC parameterization. For184
both CTM and observations, the stratosphere is defined as where O3 abundances exceed185
100 ppb (10!7 moles per mole of dry air). The 2007 climatology is an improvement over186
the 1997 climatology, but the changes also reflect the inclusion of more recent years with187
greater ozone depletion, e.g., a deeper Antarctic ozone hole in September and greater188
Arctic loss in March. From 40S to 40N, the Linoz simulation is excellent, with no obvious189
biases and errors less than 25 DU. At high latitudes, the PSC parameterization improves190
the Linoz simulations, and the ozone hole is reasonably well matched.191
To study interannual variability we use continuous ECMWF-IFS T42L40 meteorolog-192
ical fields from years 2000 through 2005. Year 2000 data are extracted from ECMWF193
IFS Cycle 23r4 model whereas the rest are extracted from Cycle 29r2 model. We find194
Cycle 29r2 generates about 20% more STE flux than does version Cycle 23r4 (see below),195
and this di!erence is much greater than the interannual variability. Thus, year 2000 me-196
teorological data are only used to spin up the experiments to approach a steady state197
before continuing with the next five years from January 2001 through December 2005,198
which are analyzed here. This five-year monthly mean climatology of total O3 column,199
plus the interannual variability defined relative to the five-year mean, are compared with200
the recent corrected Earth Probe TOMS observations based on NOAA-16 SBUV/2 ozone201
records as shown in Fig. 4. Note that the missing data for December 2005 are replaced202
with those from December 2004 for convenience. The CTM simulation with Linoz-2004203
captures the general patterns of the observed seasonal cycle and the Antarctic ozone hole204
D R A F T September 3, 2008, 12:07pm D R A F T
12 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
with its minimum below 190 DU. At mid and high northern latitudes, total O3 column205
is well simulated. In the tropics, the minimum (NDJF at 15N) are likewise matched,206
but the CTM has a spurious high (310 DU contour) in July at 10N and likewise with 270207
contour bulging equatorward to 10S in austral summer. Even worse, the circum-Antarctic208
maximum around 60S is consistently about 60 DU higher than observed. These anomalies209
do not appear in the previous publications with Linoz v1 using the 1997 and 2000-2001210
Cycle 23r4 meteorological data. Using Linoz-2004BrT reduces the total ozone error by211
20 DU confined to poleward of 60S and over the spring artic vortex. The spurious errors212
remain evident and large regardless of the chemistry used.213
Analysis of the monthly latitude-height ozone profiles from the CTM (not shown) reveals214
a deep sinking motion near the edge of the Antarctic polar vortex that persists through the215
seasons and a spurious downward shift of contours in the top model layers at 10N in July216
and 10S in January. We presume these errors stem from a poorly resolved stratosphere217
with a top lid in the middle stratosphere (2 hPa). To test this point, we acquired year218
2005 using IFS Cycle 29r2 but with much finer vertical resolution, T42L60, in which219
the whole stratosphere is resolved with layers at most 1.5 km thick from 15 to 0.5 hPa.220
Linoz-2004 with the T42L60 meteorological data corrects the worst errors seen with the221
T42L40 meteorological data as shown in Fig. 5, viz, the tropical bubble disappears and the222
circum-Antarctic high columns now are lower and closer to observations. Unfortunately,223
the T42L60 data was only available to us for year 2005, and so our analysis of interannual224
variability continues with the T42L40 data.225
Monthly anomalies in zonal-mean total O3 column for years 2001-2005 are shown in226
Fig. 4c and 4d (N.B. Contour intervals for the CTM simulation are 10 DU, but those227
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 13
for the TOMS data are only 5 DU.) In October 2002, the CTM matches the extremely228
high column anomaly over Antarctica, which was caused by a sudden warming event and229
the transport of ozone-rich air into the vortex (e.g. Simmons et al., 2005). In general230
the phases of alternating high and low anomalies are well captured by the CTM and a231
two-year QBO-like signal is evident. In spite of the coarsely resolved lower stratosphere,232
the ECMWF IFS 40-layer model produces a QBO with alternating descending easter-233
lies and westerlies in the stratosphere (not shown). The forecast model is re-initialized234
with observations every 24 hours and appears to generate a QBO pattern in the lower235
stratospheric transport. The magnitude of the modeled equatorial and SH interannual236
variability in O3 column, however, is often twice as large as observed.237
To understand the modeled interannual variability and its relation to the O3 STE flux,238
we isolate the QBO signal following the regression procedure of Randel and Wu (1996). A239
QBO time series is defined by determining the linear combination of the equatorial zonal240
wind at 20 and 40 hPa that best correlates the equatorial total O3 column interannual241
variability. This time series is then regressed against the time series of total O3 anomalies242
at all latitudes (Fig 4d). Fig. 6 shows (a) the modeled column O3 anomaly correlated243
with the QBO and (b) the residuals. The QBO signal in O3 column shows large positive244
(negative) equatorial ozone anomalies during equatorial westerlies (easterlies) as observed.245
The subtropical QBO signal is correctly out-of-phase with the equatorial signal. However,246
this signal is more confined to the subtropics than is observed (see Fig. 1 of Randel247
and Wu, 1996), and the observed 6-month phase lag between the maxima at subtropical248
and mid latitudes in the two hemispheres is absent. Also unlike earlier observations249
(e.g. Randel and Cobb 1994), the SH midlatitude QBO signal does not continue into250
D R A F T September 3, 2008, 12:07pm D R A F T
14 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
to the spring Antartic polar region but directly changes sign before 60S. This points to251
the possibility that for the coarsely resolved stratosphere of the ECMWF L40 model,252
the interaction of the annual cycles and the QBO as well as the high-latitude planetary253
waves modulated by the QBO (See Baldwin et al., 2001) are completely missing or even254
mispresented. The residuals are about the same magnitude as the QBO signal and show a255
large-scale, low-frequency, coherent structure in the SH that is roughly out of phase with256
the equator.257
4. Stratosphere-to-Troposphere Fluxes and Tropospheric Ozone
Following Hsu et al. (2005), the ozone STE flux is calculated based on the mass balance258
of a latitude-by-longitude tropospheric ozone columns. In this study, the diagnostic is259
improved by further including the first moment of the horizontal ozone flux within the260
troposphere (see equation 1 in Hsu et al., 2005) when computing this term. As a result,261
the overall noise level such as the dipole structures within the Pacific jet stream noted in262
the previous study is diminished. The average seasonal cycle of STE O3 fluxes calculated263
from the CTM simulations from years 2001-2005 (contour lines in Fig. 7) has a similar264
pattern to that published in Hsu et al (2005) for years 1997 and 2000. Also shown is265
the average zonal-mean zonal wind at 200 hPa (shaded contours). The STE maximum in266
the NH occurs just poleward of the tropospheric zonal jet, peaks during late spring and267
early summer when the zonal jet weakens, and migrates with the subtropical jet up to268
45N. In the SH, the STE maximum stays around 30S, does not migrate poleward with269
the jet in summer, but does peak during austral spring when the jet weakens. However,270
the global STE flux is on average 20 % larger than the previous estimates despite the271
fact that as discussed in Section 3, the global STE flux should be slightly reduced using272
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 15
Linoz-2004 and including the PSC parameterization. The larger magnitude is mostly due273
to the excessive decent in the Antarctic circumpolar region with the Cyc29r2 L40 fields.274
Indeed, recalculating the STE flux using the ECMWF 2005 Cyc29r2 L60 data, we find275
that the L40 SH STE flux is about 25 % too large.276
The QBOs role in modulating the STE O3 flux is derived with the same method as for277
total O3 column. STE variability attributed to the QBO (Fig. 6c) is small: peak contour278
intervals in mid-latitudes are +0.20 g m!2 yr!1 (30S in 2004) as compared with a global279
average of about 1.2 g m!2 yr!1 (i.e., 610 Tg per year). These mid-latitude STE QBO280
signals are out of phase with the mid-latitude total ozone QBO, and, not surprisingly, there281
are no QBO signals over the tropics and high latitudes where STE fluxes are negligible.282
This pattern indicates that the induced sinking part of the overturning circulation in the283
subtropics during the easterly phase of QBO creates a dynamical condition that disfavors284
the mixing of the mid-latitude, lower stratospheric ozone into the troposphere. The STE285
residual field (Fig. 6d) has much larger amplitudes than the QBO signal. In SH both286
STE and column O3 residuals show the same low-frequency variability and are positively287
correlated with STE lagging by a few months. This relationship is evident for years 2001-288
2002 (negative, blue patch) and 2003-2004 (positive, red patch). In the NH, the STE289
residuals lack coherence and have smaller amplitudes.290
Influx of O3 from the stratospheric is a principal component of the tropospheric O3291
budget, the others being in situ photochemical production and loss and surface deposition.292
The amount of tropospheric O3 ozone that can be assigned a stratospheric origin, however,293
is not well defined. There is a wide range of reported STE fluxes and various accounting294
methods for tropospheric production and loss (e.g., Prather et al., 2001; Stevenson et al.,295
D R A F T September 3, 2008, 12:07pm D R A F T
16 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
2006). In this study, we quantify the variability in tropospheric O3 caused by changes in296
STE by using a simplified, uniform chemistry in the troposphere, viz, O3 in the lowest297
600m is forced to 20 ppb. We diagnose tropospheric O3 column (adopting common usage of298
TCO for the tropospheric column ozone) hourly as the vertically integrated ozone burden299
for all CTM layers with abundances less than 100 ppb. Thus, with these simulations, the300
latitudinal and seasonal variations in TCO (color-filled contours in Fig. 8a) are driven301
by both STE flux (line contours in Fig. 8a) and the changing size of the troposphere.302
In the NH, the monthly zonal-mean TCO varies from 16 to 26 DU with lowest values303
in the tropics (where the troposphere is largest). In northern mid-latitudes, the seasonal304
peak-to-peak range is 8 DU, and although this tends to follow the troposphere mass (i.e.,305
the tropopause peaks in late summer) a large fraction appears to follow the STE flux.306
This large seasonality is driven without tropospheric chemistry. We expect the TCO to307
lag the STE by a month (i.e., the tropospheric lifetime of an STE perturbation).308
Comparing to the recently observed TCO derived from OMI and MLS measurements309
(Fig. 6 of Ziemke et al, 2006), our modelled TCO without tropospheric chemistry surpris-310
ingly matches the observed latitude-by-month pattern, much better than the comparison311
with the full chemistry CTM in Ziemke et al. Our TCO is, however, consistently biased312
low by about 8 DU over the tropics and wintertime mid-latitudes. Furthermore, we un-313
derestimate the magnitude of the buildup to maximum TCO seen at 40N in July and314
at 30S in November by 18 DU, indicating the importance of tropospheric photochemical315
net production that is not simulated here. Renormalizing the TCO pattern to the tro-316
pospheric mass shows the variation in mean tropospheric ozone abundance (ppb in Fig317
8b). Here, we see that peak abundances tend to follow the STE flux. Given that the318
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 17
tropospheric parameterization pushes to a uniform mixing ratio, this simulation provides319
a measure of the seasonality and amplitude of tropospheric O3 variability driven by STE.320
5. Ozone Depletion and STE fluxes
The reduction in STE O3 flux from the halogen-catalyzed depletion of stratospheric O3321
is evaluated with di!erent Linoz v2 models. We simulate pre-depletion O3 with Linoz-322
1979 using the meteorological data for years 2001-2005. For post-depletion, we use the323
same meteorological data and Linoz-2004. Additional experiments, Linoz-2004Br and324
Linoz-2004BrT are used to investigate the e!ect of enhanced bromine and higher PSC325
temperature threshold for post-ozone depletion. Because the additional bromine repre-326
sents a natural tropospheric source, we simulate the pre-depletion ozone, Linoz-1979Br in327
pair with the latter experiments. Note that because we have only meteorological data for328
recent years, this study cannot elucidate the role of changing transport in the observed329
depletion.330
All three Linoz-2004 variants calculate qualitatively similar patterns of column O3 de-331
pletion as observed in the merged satellite data based on TOMS/SBUV measurements332
from 1979-2000 (see Fig. 11 of Fioletov et al., 2002). The depth of overall ozone depletion333
and its seasonal di!erences, particularly in the NH, are largely underestimated unless the334
activation temperature of the PSC parameterization is raised from 195K to 199K. En-335
hanced bromine has only a small e!ect on further ozone depletion in these calculations336
compared to Salawitch et al. (2005). As noted before, we speculate that this could be the337
result of our PSC parameterized loss being independent of bromine level.338
Fig. 9 shows the latitude-by-month changes in total O3 column (DU) and STE O3339
flux (g m!2 yr!1) between Linoz-1979Br and Linoz-2004BrT for the five year average of340
D R A F T September 3, 2008, 12:07pm D R A F T
18 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
2001-2005 meteorology. We can compare the ozone depletion in Fig. 9a directly to the341
vertically integrated SAGE/sonde stratospheric ozone data over 1979-2005 (Fig. 10 of342
Randel and Wu, 2007). Overall, the simulations agree with observations, both in shape343
and magnitude. In the tropics, the di!erence of about 8 DU in total O3 column relative344
to the pre-ozone depletion is as large as seen in SAGE/sonde data (Fig. 10a of Randel345
and Wu, 2007), but disagrees with the equatorial total ozone change of about zero seen in346
the merged TOMS/SBUV data (Fig. 10b of Randel and Wu, 2007). In our simulation at347
T42L40, the top layer from 2 to 20 hPa does not accurately cover the diversity in upper348
stratospheric O3 chemistry whereby chlorine-driven depletion above 3 hPa results in more349
penetration of solar ultraviolet and hence more O3 production below. Other errors in our350
simulation include: missing the second maximum in NH ozone depletion during the fall;351
and simulating Antarctic ozone depletion to be about twice as large as the observed. The352
latter discrepancy could be reduced if there were some PSC-induced ozone loss already in353
1979, which is not modeled here.354
The latitude-month change in the STE ozone flux is quite di!erent than that in total355
ozone. For the NH STE change, the di!erence pattern follows roughly the seasonal pattern356
in Fig. 7 but with the maximum depletion in summer lagging the seasonal STE maximum357
by 2 months. The maximum change in the NH STE fluxes is about -0.3 g m!2 yr!1 (less358
than 10 % change) and is about twice that obtained when the PSC threshold is lowered359
to 195 K. The SH STE change pattern does not resemble its seasonal pattern. The360
maximum decrease, -0.7 g m!2 yr!1, occurs in the austral summer around 40S with a six361
month time-lag from the Antarctic ozone hole. Antarctic ozone, primarily in the isolated362
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 19
vortex, requires more time to propagate to the midlatitude STE exchange than its NH363
counterpart.364
Fig. 10 shows the seasonal changes in the ozone vertical profiles at NH mid latitudes365
for all three pairs of experiments. Decreases are most evident near 11-14 km in the lower366
stratosphere as the peak moves down in height from spring to summer. The annual367
average decrease in the lower stratosphere from Linoz-2004BrT pairs is about 11 %. This368
trend corresponds to about 8 % per decade for the 1979-1990 period (see Randel and Wu,369
2007) and is comparable with some estimates for a similar period ( e.g. Randel et al.,370
1999) but is much smaller than reported in Fusco and Logan (2003). With Linoz-2004Br371
pairs, the net ozone depletion is at most 5 %. The relatively uniform and weak decrease in372
the winter season is distinctively di!erent from the profiles of the other seasons and from373
observations. It might point to the importance of a trend in winter circulation (Hood and374
Soukharev, 2005) lacking in this study.375
6. Discussion and Conclusions
To separate changes in STE flux due to the ozone depletion from those due to natural376
variability, we use a regression model to fit the hemispheric monthly STE flux from the five-377
year sequence with the hemispheric mean, the seasonal harmonics and the monthly time378
series of the QBO and NAO (North Atlantic Oscillation) indices. The QBO accounts for 20379
% and the NAO for only 4 % of the NH interannual variability. The total NH interannual380
variability is about 25 Tg/yr (r.m.s), the same as the long-term change driven by ozone381
depletion (Linoz2004BrT-Linoz1979Br). Thus, as detection of a NH trend in column382
ozone is obscured by transport variability (e.g. Stolarski et al, 2006), so is any long-term383
change in STE flux. The SH STE flux shows quite di!erent modes of variability. For384
D R A F T September 3, 2008, 12:07pm D R A F T
20 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
the SH regression model, we replace NAO index with the AAO (Antarctic Oscillation)385
index. The QBO nearly captures about 45 % of the interannual variability, while the386
AAO captures negligible variance and thus does not contribute to coherent structure of387
the residuals seen in Fig. 6d. The maximum SH STE change due to ozone depletion (70388
Tg/yr) is twice as large as the interannual variability (30 Tg/year r.m.s) over the years389
2001-2005.390
The impact of changes in the STE flux on tropospheric ozone is shown with the scatter391
plot of annual, hemispheric tropospheric ozone abundances versus the mean STE fluxes392
for all 25 years of Linoz v2 simulations (Fig. 11). In both hemispheres there is a distinct393
linear correlation with slightly di!erent slopes: 0.033 ppb/Tg for the NH, and 0.028394
ppb/Tg for the SH. For given STE, the tropospheric ozone burden for NH is slightly395
lower, an indication of stronger ozone sink in the NH due to more vigorous vertical mixing396
probably driven by greater convection over continents and planetary wave activities. The397
SH STE has a much wider range due to the Antarctic ozone hole. The interannual398
variability is similar across the di!erent Linoz chemical models and thus meteorological399
variations appear to overshadow the chemical evolution of the lower stratosphere. The400
interannual variability between the NH and SH is uncorrelated (see the labelled years401
for Linoz-2004BrT). The mean di!erence due to ozone depletion between Linoz-1979Br402
and Linoz-2004BrT in the NH is 25 Tg corresponding to 1 ppb decrease for the mean403
tropospheric ozone abundance and that in the SH is 72Tg, or a 2 ppb decrease in the SH.404
This study has sought to understand and quantify how the stratosphere drives tropo-405
spheric ozone. The scientific results can be summarized as follows:406
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 21
• Linoz v2 corrects the low bias in equatorial total ozone found in Linoz v1, and with407
ECMWF IFS data our chemistry-transport model better matches the observed strato-408
spheric ozone columns.409
• Linoz is useful in diagnosing errors in stratospheric circulation for both general cir-410
culation models and assimilated winds, e.g., the change in IFS cycle to 29r2 degraded the411
EC L40 meteorology, but not the L60 version.412
• Observed interannual variability in column ozone is reasonably well modeled with413
Linoz and the EC meteorology; however, the modeled magnitude in the SH, including414
QBO, is twice as large. Other EC products such as ERA-40 with L60 resolution have also415
shown unrealistic, large SH interannual variability (Fleming et al, 2007).416
• Our best estimate for the current STE ozone flux is 290 Tg/yr in the NH and 225417
Tg/yr in the SH with interannual variability over years 2001-2005 of ± 25 Tg and ± 30418
Tg/year respectively. Enhanced STE flux can be correlated with more rapid tropospheric419
mixing and removal of ozone.420
• The STE shows negative anomalies over the mid-latitudes during the easterly phases421
of the QBO and vice versa. The QBO-induced overturning circulations over mid-latitudes422
during the easterly phase creates conditions that reduce STE.423
• The STE flux alone drives a large seasonal change in the tropospheric ozone burden424
with NH mid-latitude peak-to-peak changes of about 8 DU, mimicking summertime pho-425
tochemical production but with half the amplitude. Part of this seasonal change is due426
to increasing tropospheric air mass (i.e., rising tropopause height).427
D R A F T September 3, 2008, 12:07pm D R A F T
22 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
• When the observed column ozone depletion from 1979 to 2004 is modeled with Linoz428
v2, we predict STE reductions of about 10 %, corresponding to about 1 ppb in tropospheric429
O3 of the northern hemisphere, much less than anticipated by Fusco and Logan (2003).430
References
Baldwin, M. P., et al. (2001), The quasi-biennial oscillation, Rev. Geophys., 39(2), 179230.431
Bojkov, R. D., V. E. Fioletov (1997), Changes of the lower stratospheric ozone over Europe432
and Canada, J. Geophys. Res., 102(D1), 1337-1348, 10.1029/96JD00095.433
Cariolle, D., Lasserre-Bigory, A. and Royer, J. F. (1990), A general circulation model sim-434
ulation of the springtime antarctic ozone and its impact on mid-latitudes. J. Geophys.435
Res., 95, 1883-1898.436
Cariolle, D. and H. Teyssedre (2007), A revised linear ozone photochemistry parameter-437
ization for use in transport and general circulation models: multi-annual simulations.438
Atmos. Phys. Disscuss, 7, 1655-1697, 2007.439
Eyring, V., D. E. Kinnison, and T. G. Shepherd (2005), Overview of planned coupled440
chemistry-climate simulations to support upcoming ozone and climate assessments,441
SPARC Newsletter No. 25, 11-17.442
Fleming, E. L., C. H. Jackman, D. K. Weisenstein, and M. K. W. Ko (2007), The impact443
of interannual variability on multidecadal total ozone simulations, J. Geophys. Res.,444
112, D10310, doi:10.1029/2006JD007953.445
Fusco A. C., J. A. Logan (2003), Analysis of 19701995 trends in tropospheric ozone at446
Northern Hemisphere midlatitudes with the GEOS-CHEM model, J. Geophys. Res.,447
108 (D15), 4449, doi:10.1029/2002JD002742.448
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 23
Hsu J., M. J. Prather, O. Wild (2005), Diagnosing the stratosphere-to-troposphere449
flux of ozone in a chemistry transport model, J. Geophys. Res., 110, D19305,450
doi:10.1029/2005JD006045.451
IUPAC (2004), International Union of Pure and Applied Chemistry, Subcommittee on Gas452
Kinetic Data Evaluation, http://www.iupac-kinetic.ch.cam.ac.uk/, Datasheets down-453
loaded 18 May 2004.454
McPeters, R. D., G. J. Labow, and B. J. Johnson (1997), A satellite-derived ozone clima-455
tology for balloonsonde estimation of total column ozone, J. Geophys. Res., 102(D7),456
88758886.457
McPeters R. D., G. J. Labow, J. A. Logan (2007), Ozone climatological profiles for satellite458
retrieval algorithms, J. Geophys. Res., 112, D05308, doi:10.1029/2005JD006823.459
McLinden, C.A., S. C. Olsen and B.J. Hannegan, O. Wild and M.J. Prather (2000),460
Stratosphere ozone in 3-D models: A simple chemistry and the cross-tropopause flux,461
J. Geophys. Res., 105(D11), 14653-14666, 10.1029/2000JD900124.462
Michelsen, H. A., G. L. Manney, M. R. Gunson, R. Zander (1998), Correlations of strato-463
spheric abundances of NOy, O 3, N 2 O, and CH 4 derived from ATMOS measurements,464
J. Geophys. Res., 103(D21), 28347-28360, 10.1029/98JD02850.465
Michelsen, H. A., G. L. Manney, M. R. Gunson, C. P. Rinsland, R. Zander (1998), Correla-466
tions of stratospheric abundances of CH4 and N2O derived from ATMOS measurements,467
Geophys. Res. Lett., 25(15), 2777-2780, 10.1029/98GL01977.468
Nassar R., P. F. Bernath, C. D. Boone, G. L. Manney, S. D. McLeod, C. P. Rinsland, R.469
Skelton, K. A. Walker (2005), Stratospheric abundances of water and methane based on470
ACE-FTS measurements, Geophys. Res. Lett., 32, L15S04, doi:10.1029/2005GL022383.471
D R A F T September 3, 2008, 12:07pm D R A F T
24 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
Nagatani, R. M., and J. E. Rosenfield, Temperature (1993), net heating and circulation,472
in The Atmospheric E!ects of Stratospheric Aircraft: Report of the 1992 Models and473
Measurements Workshop, NASA Reft Publ. 1292, edited by M. J. Prather and E. E.474
Remsberg, pp. A1-A47.475
Prather, M. (1992), Catastrophic loss of stratospheric ozone in dense volcanic clouds, J.476
Geophys. Res., 97(D9), 10,18710,191.477
Prather, Michael and E Remsburg (1993), The atmospheric e!ects of stratospheric air-478
craft. Report of the 1992 models and measurements Workshop. NASA Publ 1292.479
Prather, M., D. Ehhalt, F. Dentener, R. G. Derwent, E. Dlugokencky, E. Holland, I. S. A.480
Isaksen, J. Katima, V. Kirchho!, P. Matson, P. M. Midgley, and M. Wang, Chapter 4481
(2001), Atmospheric Chemistry and Greenhouse Gases, in Climate Change 2001: The482
Scientific Basis, J.T. Houghton et al., eds., Cambridge U. Press, pp. 239-287.483
Randel, W. J., and J. B. Cobb (1994), Coherent variations of monthly mean total ozone484
and lower stratospheric temperature, J. Geophys. Res., 99(D3), 54335447.485
Randel, W.J., B.A. Boville, J.C. Gille, P.L. Bailey, S.T. Massie, J.B. Kumer, J.L. Mer-486
genthaler, and A.E. Rochem (1994), Simulation of Stratospheric N2O in the NCAR487
CCM2: Comparison with CLAES Data and Global Budget Analyses. J. Atmos. Sci.,488
51, 2834-2845.489
Randel, W.J., and F. Wu, 1996: Isolation of the Ozone QBO in SAGE II Data by Singular-490
Value Decomposition. J. Atmos. Sci., 53, 2546-2559.491
Randel W. J., R.S. Stolarski, D.M. Cunnold, J.A. Logan, M. J. Newchurch, and J. M.492
Zawodny, 1999: Trends in the Vertical Distribution of Ozone, Science 285 (5434), 689,493
DOI:10.1126/science.285.5434.1689.494
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 25
Randel W. J., F. Wu (2007), A stratospheric ozone profile data set for 19792005: Variabil-495
ity, trends, and comparisons with column ozone data, J. Geophys. Res., 112, D06313,496
doi:10.1029/2006JD007339.497
Salawitch R. J., D. K. Weisenstein, L. J. Kovalenko, C. E. Sioris, P. O. Wennberg, K.498
Chance, M. K. W. Ko, C. A. McLinden (2005), Sensitivity of ozone to bromine in the499
lower stratosphere, Geophys. Res. Lett., 32, L05811, doi:10.1029/2004GL021504.500
Sander et al. (2006), Chemical kinetics and photochemical data for use in atmospheric501
studies, Evaluation Number 15. JPL Publication 06-2., Jet Propul. Lab., Pasadena,502
Calif.503
Schoeberl, M.A. et al. (1993), The evolution of ClO and NO along air parcel trajectories,504
Geophys. Res. Lett., 20(22), 2511-2514505
Simmons A., M. Hortal, G. Kelly, A. McNally, A. Untch and S. Uppala (2005), ECMWF506
analyses and forecasts of stratopsheric winter polar vortex breakup: September 2002 in507
the southern hemisphere and related events. J. Atmos. Sci., 62, 668-689.508
Stevenson, D. S., et al. (2006), Multimodel ensemble simulations of present-day and near-509
future tropospheric ozone, J. Geophys. Res., 111, D08301, doi:10.1029/2005JD006338.510
Stolarski, R.S., A.R. Douglass, S. Steenrod, and S. Pawson (2006), Trends in Stratospheric511
Ozone: Lessons Learned from a 3D Chemical Transport Model. J. Atmos. Sci., 63,512
10281041.513
Strahan, S. E. (1999), Climatologies of lower stratospheric NOy and O3 and correlations514
with N2O based on in situ observations, J. Geophys. Res., 104(D23), 30463-30480,515
10.1029/1999JD900775.516
D R A F T September 3, 2008, 12:07pm D R A F T
26 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
Sudo, K., M. Takahashi, and H. Akimoto (2003), Future changes in stratosphere-517
troposphere exchange and their impacts on future tropospheric ozone simulations, Geo-518
phys. Res. Lett., 30(24), 2256, doi:10.1029/2003GL018526.519
Thornton, B. F. et al. (2005), Variability of active chlorine in the lowermost Arctic strato-520
sphere, J. Geophys. Res., 110(D22), D22304.521
Thuillier, G., L. Floyd, T. N. Woods, R. Cebula, E. Hilsenrath, M. Herse, D. Labs (2004),522
Solar Irradiance Spectra for Two Solar Activity Levels, Adv. Space Res., 34, 256-261.523
Wamsley, P. R., et al. (1998), Distribution of halon-1211 in the upper troposphere and524
lower stratosphere and the 1994 total bromine budget, J. Geophys. Res., 103(D1), 1513-525
1526.526
Wild, O., J. K. Sundet, M. J. Prather, I. S. A. Isaksen, H. Akimoto, E. V. Browell,527
and S. J. Oltmans (2003), Chemical transport model ozone simulations for spring528
2001 over the western Pacific: Comparisons with TRACE-P lidar, ozonesondes, and529
Total Ozone Mapping Spectrometer columns, J. Geophys. Res., 108(D21), 8826,530
doi:10.1029/2002JD003283.531
WMO (World Meteorological Organization), Scientific Assessment of Ozone Depletion:532
2006, Global Ozone Research and Monitoring ProjectReport No. 50, 572 pp., Geneva,533
Switzerland, 2007.534
Woodbridge, E. L., et al. (1995), Estimates of total organic and inorganic chlorine in the535
lower stratosphere from in situ and flask measurements during AASE II, J. Geophys.536
Res., 100(D2), 3057-3064.537
Ziemke, J. R., S. Chandra, B. N. Duncan, L. Froidevaux, P. K. Bhartia, P. F. Levelt, and538
J. W. Waters (2006), Tropospheric ozone determined from Aura OMI and MLS: Evalu-539
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 27
ation of measurements and comparison with the Global Modeling Initiative’s Chemical540
Transport Model, J. Geophys. Res., 111, D19303, doi:10.1029/2006JD007089.541
D R A F T September 3, 2008, 12:07pm D R A F T
28 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
0.1
0.20.30.4
0.50.6
0.70.80.9
0.1
0.2
0.20.20.3 0.3
0.3
0.4
0.4
0.5
0.5
0.6
0.60.7
0.70.8
0.8 0.9
0.10.2
0.30.4
0.50.6
0.70.8 0.9
0.10.2 0.30.4
0.50.6
0.7
0.80.9
1 1
0.2
0.30.40.50.60.7
0.80.9
Latitude
1.1
1.21.
31.41.5
1.6
1.7
N2O (318 ppb) CH4 (1.78 ppm) H2O (3.65 ppm)
NOy (19.4 ppb)
Cly (3.44 ppb) Bry (15.6 ppt)
!50 0 50!50 0 50 !50 0 50
Hei
ght (
km)
0
10
20
30
40
50
0
10
20
30
40
50
Figure 1. The Linoz climatologies of trace gases for January. For N2O and CH4 the
patterns are normalized relative to their mean tropospheric abundances; and for H2O, to
the tropopause value. The trace gas families (NOy, Cly, Bry) are normalized relative
to their maximum values (in the upper stratosphere). For year 2004, the normalization
values are noted.
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 29
!1.0
!0.5
0.0
0.5
1.0
x106
P!L
(cm
!3s!1
)
10 20 30 40 50 60
4
5
6
7
Height (km)
!10
!8
!6
!4
!2
0
x104
d(P!
L)/d
T (c
m!3
s!1
K!1
)
0
2
4
6
8
10
12
x105
d(P
!L)/
dO3co
l (cm
!3 s
!1 D
U!1
)
JPL97!S3
JPL97!S2
JPL97!S1
JPL00!S1
JPL02!S1
JPL06!S1
P!L
d(P!L)/dO3
d(P!L)/dT
d(P!L)/d(O col)3
!2
8!
log
d
(P!L
)/dO
(
s )!1
10
3
Figure 2. The sensitivity of Linoz terms (Equation 1) to di!erent radiation and chemical
updates using the standard ATMOS-profile on May 31 at 30N as the basic state. See the
text for details.
D R A F T September 3, 2008, 12:07pm D R A F T
30 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
200
250
300
350
400
MAR
DU
150
200
250
300
350
400 JUN
DU
150
200
250
300
350
400 SEP
DU
!50 0 50
150
200
250
300
350
400DEC
latitude
DU
450Linoz-2004 w/o PSCObs - 1997Obs - 2007
Figure 3. Stratospheric O3 columns (DU) as a function of latitude for March, June,
September and December. The stratospheric column is integrated over the atmosphere
where O3 > 100 ppb. The four di!erent profiles are: McPeters et al. (1997) climatology
(black line with o); McPeters et al. (2007) climatology (black line with +); Linoz-2004
in the UCI CTM with the 1997 ECMWF IFS met data (red solid line); the same Linoz
simulation without PSC parameterization (red dashed).
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 31
170190210230 250
250
250270270
270
270
290290
290
290290
310
310
330
330
350350
310
31037
039
0
330
310410
350
330
(a) EP TOMS Clim
J F M A M J J A S O N D J
−60
−30
0
30
60
−40−35−20−15
−15
−15−10 −10
−10
−10
−10
−10
−5−5
−5
−5 −5−5
−5 −5
−5−5 −5
−5−5
5
5
5
5
5 55
55
5
5
5 55
5
5
5
5
5
5
55
10
10
10
10
10
10
15
15
15
2025303540
(c) EP TOMS Ano
J J J J J
−60
−30
0
30
60
230
250250 270
270
270
270
290
290
290
290
310
310
310
310
330
330
350
350
330350370390
410
370
370
330
430
350
450
250
370
310
390
390
(b) UCI CTM/ECMWF−IFS Clim
J F M A M J J A S O N D J
−60
−30
0
30
60
−50
−40
−40
−40−30 −30
−30
−30
−30−30
−20 −20
−20
−20
−20
−20
−20
−20
−10
−10
−10
−10
−10
−10
−10
−10
−10
−10
−10
−10
−10
−10
−10
−10
−10
−10
10
10
10
10 10
10
1010
10
10
1010
10
10
1010
10
10
20
20 20
20
20
20
20 20
30
30 30
30 30
30
40
40
40
40
50
50 607080
(d) UCI CTM/ECMWF−IFS Ano
J J J J J
−60
−30
0
30
60
Figure 4. Five-year (2001-2005) monthly zonal-mean climatology of total O3 column
(DU) as a function of latitude and for (a) corrected Earth Probe TOMS and (b) Linoz-
2004 CTM simulation using Oslo/EC meteorological data at T42L40 resolution from IFS
Cycle 29r2. The anomalies reltive to this climatology from Jan 2001 to Dec 2005 are
shown for (c) TOMS and (d) CTM.
D R A F T September 3, 2008, 12:07pm D R A F T
32 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
175200
225
250
250250
250
250
250
250
250
275
275275
275
275
275
275
275275
300
300300300
300
300 300
300
300
325 325
325
325
325
325
325
325325
350
350
350
350
350350
375
375 325
325
400400
375
375
375
350
300
300
425
450
400
400
375
22535
0
225
toatal cloumn ozone (DU) 2005T42L40
J F M A M J J A S O N D J
−80
−60
−40
−20
0
20
40
60
80
175
200
225
250
250250
250
250 250
250
250
250
275
275275
300300
300
275275
275
275
275
325
325
325
325325
300 300
300
300
300
300
350
350
325
325
325
325
325
375
375
400
400
350
350
350350
425350
375375
375
450
325
375
475
400
toatal cloumn ozone (DU) 2005T42L60
J F M A M J J A S O N D J
−80
−60
−40
−20
0
20
40
60
80
Figure 5. Year 2005 monthly zonal mean total column O3 (DU) simulated with Linoz-
2004 for (a) T42L40 CTM and (b) T42L60 CTM.
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 33
time
Latit
ude
(a) total ozone QBO fit (DU)
J J J J J
−60
−30
0
30
60
(b) total ozone residual (DU)
time
Latit
ude
J J J J J
−60
−30
0
30
60
time
latit
ude
(c) STE QBO fit (g/m2/year)
J J J J J
−60
−30
0
30
60
(d) ozone STE residual (g/m2/year)
time
latit
ude
J J J J J
−60
−30
0
30
60
Figure 6. Monthly zonal mean anomalies in the modeled total O3 column split into
(a) QBO signal and (b) residuals, taken from the 2001-2005 simulation shown in Fig. 4d.
Contour intervals are +5, +10, +15, DU (solid red) and -5, -10, -15, (dashed blue).
Corresponding monthly zonal mean anomalies in the STE O3 fluxes are shown for (c)
QBO signal with contour intervals of ±0.05 g m!2 yr!1 and (d) the residuals with contour
intervals of ±0.10 g m!2 yr!1. Zero contour lines are omitted.
D R A F T September 3, 2008, 12:07pm D R A F T
34 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
Latit
ude
O3 STE flux (g/m2/year) and zonal mean zonal wind (m/sec)
0 0
1 1
1
1
1 1
1
11
1
11
1
1
1 1
2
2
22
2
2
2
2
33
3
0.5
0.5
0.5 0.5
0.5
0.50.5
0.5 0.5
0.5
0.5 0.5
0.50.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.51.5
1.51.5
2.5
2.52.5
2.5
3.5
J F M A M J J A S O N D J
−70
−50
−30
−10
10
30
50
70
−10
−5
0
5
10
15
20
25
30
35
40
Figure 7. Latitude by month average STE O3 fluxes (white-line contours at 0 to +3.5
g m!2 yr!1) from UCI CTM with Linoz-2004 driven by ECMWF IFS T42L40 2001-2005
met data. Zonal-mean zonal wind at 200 hPa (grey-scale contours at -5, +5, +15, +25,
+35 ms!1) from the same met data.
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 35
Latit
ude
STE flux (g/m2/year) and TOC (in DU)
0 0
1 1
1
1
1 11
11 1
11 1
1
11
2
2
22
2
2
2
23 33
0.5 0.5
0 5 0.5
0.50.5 0.5
0.5 0.50.5
0.5 0.5
0.50.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.51.5
1.5 1.5
2.5
2.52.5
2.5
3.5
J F M A M J J A S O N D J
−70
−50
−30
−10
10
30
50
70
10
12
14
16
18
20
22
24
26La
titud
e
STE flux (g/m2/year) and Trop. O3 Concentration (ppb)
0 0
1 1
1
1
1 1
1
1
11
11 1
1
11
2
2
2
2
2
2
2
2 3 33
0.5 0.50.5 0.5
0.5 0.50.5
0.50.5 0.5
0.5 0.5
0.50.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.51.5
1.5 1.5
2.52.5
2.5
2.5
3.5
J F M A M J J A S O N D J
−70
−50
−30
−10
10
30
50
70
20
25
30
35
40
45
Figure 8. (a) Latitude by month average STE O3 fluxes (see Fig 7) on top of simulated
tropospheric column ozone (TCO, color-filled contours at 12, 14, 16, 18, 20, 22, 24 and
26 DU). (b) Monthly zonal-mean tropospheric O3 abundance (ppb).
D R A F T September 3, 2008, 12:07pm D R A F T
36 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
−160−140−120
−120
−100 −100−80
−80−60
−60
−40
−40
−40
−40 −40−3
0
−30
−30
−30−30
−30
−20−20
−20
−20−20
−20
−16−16
−16
−16−16
−16
−12 −12
−12
−12
−12 −12 −12
−8−8−8
−8
−8 −8 −8
J F M A M J J A S O N D J
−60
−40
−20
0
20
40
60
−0.7
−0.7−0.5
−0.5−0.5
−0.5
−0.3 −0.3−0.3
−0.3−0.3
−0.3
−0.3
−0.3
−0.3
−0.1 −0.1−0.1
−0.1
−0.1
−0.1
−0.1
−0.1
−0.1
−0.1 −0.1−0.1
−0.1
J F M A M J J A S O N D J
−60
−40
−20
0
20
40
60
Figure 9. (a) Monthly zonal-mean di!erences in total O3 column (DU) for Linoz-
2004BrT minus Linoz-1979Br calculated with the same 5-year met data. Contour intervals
are 8, 12, 16, 20, 30, 40, DU. (b) Di!erences in STE O3 flux for the same simulation.
Contour intervals are 0.1, 0.2, 0.3, 0.4, and 0.5 g m-2 yr-1.
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 37
−15 −10 −5 05
10
15
20
heig
ht (k
m)
%
DJF: ozone change
−15 −10 −5 05
10
15
20
heig
ht (k
m)
%
MAM: ozone change
−15 −10 −5 05
10
15
20
heig
ht (k
m)
%
JJA: ozone change
−15 −10 −5 05
10
15
20
heig
ht (k
m)
%
SON: ozone change
Figure 10. Seasonal profile changes in O3 abundance (%) over the lower strato-
sphere and upper troposphere at 40N-50N for 2004 relative to 1979. Base calculations
use Linoz-1979 and Linoz-1979Br and perturbation calculations use Linoz-2004 (crosses),
Linoz-2004Br (circles) and Linoz-2004BrT (triangles). All calculations use the same me-
teorological data.
D R A F T September 3, 2008, 12:07pm D R A F T
38 HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE
200 250 300 350 40030
31
32
33
34
35
36
O3 STE flux (Tg)
Trop
O3 (p
pb)
1
2
3
4
5
12 3
4
5
1
2
3
4
5
12 3
4
5
Linoz−1979Linoz−2004Linoz−1979BrLinoz−2004BrLinoz−2004BrT
Figure 11. Mean tropospheric O3 (ppb) vs. STE O3 flux (Tg/yr) by hemisphere for
five di!erent Linoz models: -1979 (red crosses), -1979Br (black squares), -2004 (green
left-triangles), -2004Br (blue circles), and -2004BrT (blue asterisks). Values for the five
meteorological years, 2001-2005, are shown as 1-5 and only labeled for Linoz-2004BrT.
The upper dashed-dot line is a fit to SH data with the slope of 0.028 (ppb/Tg), and the
lower dashed blue line, to NH data with the slope of 0.033 (ppb/Tg).
D R A F T September 3, 2008, 12:07pm D R A F T
HSU AND PRATHER: STRATOSPHERIC VARIABILITY AND TROPOSPHERIC OZONE 39
Species abundances\Linoz v2 Linoz-1979 Linoz-2004 Linoz-1979Br Linoz-2004Br Linoz-2004BrT
N2O (ppbv) 300.4 318.4 300.4 318.4 318.4
NOy (ppbv) 18.2 19.4 18.2 19.4 19.4
Cly (pptv) 2242 3437 2242. 3437 3437
Bry (pptv) 8.7 15.6 14.7 21.6 21.6
CH4 (ppbv) 1555 1777 1555 1777 1777
H2O (ppmv) 3.65 3.65 3.65 3.65 3.65
PSC activa. Temp (K) 195 195 195 195 199
Table 1. Prescribed abundances of long-lived species and activation temperatures for
the PSC parameterization used for deriving the 5 Linoz tables indicated in the column
headings. See text for details.
D R A F T September 3, 2008, 12:07pm D R A F T