Dussaillant, 2007

7/27/2019 Dussaillant, 2007

1/13

Desalination 246 (2009) 202214

Water level fluctuations in a coastal lagoon: El Yali Ramsar wetland, Chile

Alejandro Dussaillant a , Pablo Galdames b , Chi-Le Sun ba Department of Civil Engineering and EULA Environmental Center, Universidad de Concepcio n, Chile

b Department of Civil and Environmental Engineering, Pontificia Universidad Cato lica, ChileTel: + 56 41 2204320; Fax:+ 56 2 354 5850; email: [email protected]

Received 19 September 2007; revised 14 February 2008; accepted 07 March 2008

Abstract

El Yali coastal reserve is the most important wetland complex in Mediterranean climate central Chile, espe-cially due to the native and foreign bird fauna which arrives here periodically. The coastal lagoon, part of a micro-tidal estuary (1.2 m tidal range), is a shallow (< 1 m depth) dynamic system and unique site of coexistence of northern halophyte and southern palustrian riparian vegetation. This study identifies and quantifies the effectof forcing variables in the lagoon water level over 1 year of data collection. Transects of piezometers withlevel sensors were installed between the coastal lagoon and the sea. Monthly water quality data were collected.During the winter rainy season, the lagoon connects with the sea via an ephemeral tidal inlet, producing noticeabledaily variations in the water level, up to 80-cm depending on the tides. In contrast, during the season when bar closure of the inlet disconnects the lagoon from the sea, the lagoon level is very stable and only decreases veryslowly due to evaporation, which also makes the system hypersaline. During the connection phase, analyses usinggeneral pattern, spectral and Fourier analysis of the sea-vs. lagoon-level signals show that two temporal scale hier-archies are relevant: monthly (due to moon cycles) and daily (due to tidal cycles every 12.5 and 24.2-h). A simplediffusion numerical model simulated the water table trends well for the sand bar between the lagoon and the sea,supporting the main effect of the sea level on the lagoon water levels.

Keywords : Coastal lagoon; Groundwater; Semi arid area; Riparian; Eutrophication; Fourier

1. Introduction

Wetlands are considered among the most pro-ductive ecosystems and key service providers tohumankind [18]. Coastal lagoons in particular,

being in the boundary or ecotone between theland and the sea, are delicate and dynamic eco-systems, exposed to frequent fluctuations and alterations. Hydrodynamic processes includeforcings by meteorology, tides, winds and spa-tial/temporal variability of salinity and tempera-ture [1,58]. In a coastal lagoon, the spatialCorresponding author.

Presented at Multi Functions of Wetland Systems, International Conference of Multiple Roles of Wetlands, June 2629, 2007, Legnaro (Padova) Italy

doi:10.1016/j.desal.2008.03.0530011-9164/09/$ See front matter 2009 Published by Elsevier B.V.


2/13

distribution of salinity usually affects biota and hydraulics of the wetland. On the other hand, hal-ophyte vegetation stabilizes dunes, increaseshydraulic roughness, promotes sediment deposi-tion and inhibits dune propagation due to wind suspension [6]. Due to salinity and flooding gra-dients, a vegetation zonation is produced, favour-ing high biodiversity [6,9]. The tide influencesthe wetland through rising of sea level and pro-motes a wave of groundwater-level oscillationthat diffuses through the system, slowly if through the subsurface, or fast if surface water mediated [8,10].

Studies available on coastal wetlands haveshown interesting connections between the

hydrology and the ecology; in particular, rele-vant studies associated to our investigationinclude those at other similar Mediterraneanregions such as eastern Spain [7,11], Portugal[12], Italy [8], Greece [5], Australia [13] and Mexico [14].

The most important wetland complex in cen-tral Chile is El Yali. This area presents a highseasonal dynamism (wintersummer), plusinterannual cycles due to El Nin o SouthernOscillation (ENSO), and is the only wetland area in the neotropics protected by the Ramsar Convention [15]. It is key habitat for migratory

birds more than 115 species, which representmore than 25% of native Chilean bird speciesdiversity [16]. Plant biodiversity is not veryhigh, but it is a unique ecosystem being a north-ern distribution limit for typical palustrian plantspecies (e.g. Spartina densiflora ) and southerndistribution limit for typical halophyte speciesfrom northern Chile (e.g. Salicornia fruticosa ).

The coastal lagoon (Fig. 1) attracts particular attention: it is one of the few water bodies thatis encapsulated totally inside the reserve; itattracts a mixed diversity of sea and inland spe-cies; and is being threatened by human activitiesdirectly (cattle trespassing, stream flow regimealterations, tourist settlements) and indirectly(high-income crop production, urbanization,

wastewater discharges to streams and soils).Though unique, several other wetlands in thecoastal central region of Chile share a similar context of coastal eroded watersheds undergoingrapid land use changes that may affect coastalecosystem productivity [17], dynamics (seasonalstreamflow opens outlet bars in winter months

but disconnects the estuary for most of theyear), and ecological importance (habitat for migratory bird, fish and other species). Thearea has been the subject of some recent ecolog-ical studies concerning mainly riparian vegeta-tion distribution [18].

Our general goal is to understand lagoonhydrology and connections to ecology in this a

unique site, as well as its use as a template for other similar functioning systems in centralChile. This investigation thus represents a firststep in this direction, seeking to gain further understanding of the relationship between seaand lagoon water level, as mediated by thelagoonocean connectedness through the Yalistream mouth. We use a combination of field

Pacific Ocean

Coastal lagoonEl Yali

Yali Stream

1 Km (approx.)

N

Fig. 1. Yali coastal lagoon (CONAF 1998).

203 A. Dussaillant et al. / Desalination 246 (2009) 202 214


3/13

observations, time series analysis and numericalmodelling of the lagoonocean-level interactionto demonstrate the main influence of sea level onlagoons water level and possible connections toriparian groundwater levels.

2. Methods

2.1. Study site

The climate in the area is Mediterranean withoceanic influence and marked seasonality.Autumnwinter (MayAugust) rainfall repre-sents 90% of the annual rainfall, followed by along period of seven dry months [19]. This dic-

tates the Yali stream flow regime, with an esti-mated average (there is no gauging station)annual flow of 1.2-m 3 /s, mostly during half of the year with negligible flows during the dry sea-son. Based on local data (station 10-km fromlagoon), mean annual rainfall is 481-mm,while potential evaporation in the area has

been estimated as 1500-mm [20, 21].The coastal lagoon has an elongated shape

parallel to the beach, being formed by the inter- play of Yali streamflow and ocean influencemediated by a mouth that closes in the dry sea-son (Fig. 1). During dry months, stream flow isnot enough to prevent a bar closure of the outletto the ocean, a period we will refer to asdisconnected. Following the classification of Cooper [22], the lagoon would be a non-

perched closed estuary, with no high berm(low sand bar), sporadic surface channel connec-tion to the ocean, dissipative (low gradient) pro-file beaches and wide surf zones, with salt marsh

vegetation. The lagoon is very shallow (less than0.5-m average depth), which together withwindy conditions provides a very well mixed water column [23]. The surface channel typi-cally is also very shallow, with only a few cmdepth, and several meters wide. Disconnected

periods provide very stable water area and volume, and thus a more stable habitat [22].

Due to high evaporation potential, lagoonwater is very saline, particularly as the dry

period advances (disconnected phase). Never-theless, we note that the Yali lagoon is differentfrom the South African systems described byCooper [22], since the outlet is seasonal [23].

We have postulated [23,24] that the lagoonhydrologic balance is a function of the followingelements with different importance dependingon the season (conceptual model in Fig. 2):

dsdt

R E QY Q

D

Q s+ QGS + QSS QOS 1

where the terms are (an asterisk

indicates thatthe flux occurs only during winter, i.e. the rainyseason) the following: S , lagoon water storage;

R , rainfall; E , lagoon evaporation; surface

inflow, QY from Yali stream, and Q D

, drainageditch flows, which mainly occur during stormsdue to soil saturation [24]; QS , groundwater dis-charge from the upland aquifer; QGS , subsurfaceflows from/to the ocean due to sea level vs.lagoon water level difference; and surfacewater connection with the ocean QSS

, due totide influence through open inlet part of theyear (together with QOS , occasional stormsurge over washing, in parentheses). Note thatwe do not include stormwater overland flowsince the soil is highly permeable sand, and that we are not reporting estimates for manyof the water budget fluxes mentioned above,due to lack of data and monitoring availability;moreover, it is not the objective of this paper

but of future work.

2.2. Monitoring methods

Due to the lagoon configuration, one can rec-ognize two distinct groundwater zones: oneupslope from the lagoon, where groundwater flows from the watershed aquifer and dischargesinto the lagoon [20,21]; and the other being the

A. Dussaillant et al. / Desalination 246 (2009) 202214204


4/13

sand bar between lagoon and ocean, includingthe beach [23]. For groundwater monitoring,three piezometer transects were installed in

20042005, perpendicular to the beachlagoonlongitudinal axis (Fig. 2), so as to investigateaquifer connection (23 piezometers in south-east/upslope side) but principally, at least atthis stage, sealagoon connectivity (3 piezome-ters in northwest/sea side). Piezometer lengthswere 1 to 2-m deep. Water levels were initiallymeasured in seasonal scouting campaigns, start-ing in January 2006 using TruTrack TM transducer sensors ( + 1-cm precision) in 10-min intervals[23]. Sea level data in hourly intervals were pro-vided by the Navy Oceanographic Station in SanAntonio, 20 km from the site. Typical tideamplitude range is 1.2 m (i.e. Yali streammouth is a microtidal estuary). Daily rainfalldata are recorded at a station 10 km from thelagoon, assumed to be the same due to Chileanfrontal-type rainstorms and same altitude.

Wind data (hourly intensity and direction from9 am to 6 pm) are registered at another station15-km from Yali wetland, again at similar alti-

tude and exposition. A summary of the dataused is presented in Table 1.Saturated hydraulic conductivity was esti-

mated using constant-head permeameter testson samples collected when piezometers wereinstalled, and complemented with measurementstaken later using a tensioinfiltrometer [23,24].Lagoon bottom sediment permeability was esti-mated using seepage meters installed in fivelocations and using average measured heads inthe 1520-m vicinity as an approximation.

Since there is concern of anthropogeniceutrophication from the viewpoint of naturalresource managers ( Corporacio n Nacional

Forestal [CONAF]), who provided field accessand support, and there might be interest incomparing to other similar systems worldwide,we provide some summarized reference data,

CO399,75

CO299,67

CO198,74

CP199,53

CP299,28

CP399,63

CP499,75

CP599,94

44,247,8 159,927,039,6 15,513,8 65,4

SEA

LAGOON

SEALAGOON

QOS

RE

QS

QY

QSS QGS

QD Piezometer transects

(Seepage meters)

Fig. 2. Conceptual diagram of hydrological fluxes: plain view (top, piezometer transects shown) and cross section viewof central piezometer transect (bottom, seepage meter location shown, and with central transect coding: CP for betweenthe sea and the lagoon and CO for upslope lower numbers given to piezometers closer to the lagoon, with relativeheight of transducer under the respective piezometer code, and horizontal spacing, in cursive, at the figure bottom) all numbers in meters.



5/13

which will be published in full in a future paper based on synthesis of ongoing work [23,24]. Dueto the strong dependence of lagoon level on tideswhile connected to the sea, water quality measure-ments have been done monthly on full moon days,when the maximum tidal amplitude differencesare registered and therefore higher fluxes areassumed to occur. Measurements included dis-solved oxygen, pH and conductivity with on-sitesensors; and nutrients from samples stored at48C, filtered and analyzed with spectrophotometer methods. Table 2 summarizes the methods.

2.3. Data analysis

To analyse the relation between sea level and lagoon depth time series, we used several

methods, including: (1) typical average patternrepetition and residual analysis [25], to identifyaverage characteristics of rising and recessingcyclic periods; (2) frequency spectrums, to iden-tify most important frequencies of processesaffecting water level, in connected and discon-nected phases; (3) Fourier analysis, to relate seato lagoon water level, through a typical impulseresponse function approach [26]; and (4) a simpli-fied numerical model based on Darcys equationapplied to an unconfined aquifer, assuming onlylateral flow (Dupuit approximation), thus reduc-ing to a diffusion-type equation:

h t

D 2 h x2

w p

2

Table 1Summary of data used

Type of Data Unit Resolution Precision Range

Water level m 10-min mean + 1 cm JanNov 2006Sea level m Hourly mean + 1 cm JanNov 2006Rainfall mm Daily + 0.25 mm JanSep 2006Wind m/s Hourly (918 h) + 0.1 m/s JanSep 2006

Table 2Water quality equipment and methods used

Instrument (brand) Parameter Precision Range

WT-HR (TruTrack) Temperature + 0.18C 308C to + 70 8CSension5 (HACH) { Conductivity { + 0.5% 0.00 mS/cm to 199.9 mS/cm

Salinity + 0.1% 0 to 42 %TDS { + 0.5% 0.00 mg/L to 50g/L

Triode pH Electrode (Thermo Orion) pH + 0.1 0 to 14DO 100 (Oakton) OD + 1.5% 0 to 20.00 mg/L

0 to 200.0%Spectrophotometer DR/2010 (HACH) N-NO 3

+ 0.10 mg N-NO 3 /L 0.0 to 4.5mgN-NO 3

/LP-PO 4 + 0.01 mg PO 4

3 /L 0.0 to 2.5 m g P-PO 43 /L

(+ 0.003 mg P-PO 4 /L) 0.0 to 0.82 mg P-PO 4 /L

{Corrects for temperature for samples between 2 and 35 8C.{Scale adjusts automatically according to the measurement. From the respective measurement scale. Corrects for temperature for samples between 0 and 50 8C and for salinity between 0.0 and 50.0 % .



6/13

D ka p

3

where h is hydraulic head [L], measured; w is therainfall rate [L/T], measured; D is diffusioncoefficient, assumed constant, given by Eq. 3,with k saturated hydraulic conductivity [L/T],estimated; a is the aquifers thickness [L](approximately 100 m [27]); and p is soil poros-ity [L 3 /L3 ], measured. This partial differentialequation was solved for h with a fully implicitmethod, using hourly lagoon-water-level and sea-level data as boundary conditions, and theinitial condition being an interpolation between

both.

3. Results and discussion

3.1. Soil hydraulics

Soils in the vicinity of the lagoon are coarseand relatively homogeneous ( D50 = 0.18-mm,

porosity = 0.43), primarily sand with some gravel.K s, soil saturated hydraulic conductivity, wasmeasured both in the field (using a tensioinfil-trometer) and in the lab (permeameter measure-ments from soil samples taken from piezometer installation). It ranged from 0.80 to 13-m-d 1

on the upslope (SE) side; the sea (NW) side K sranged from 0.53 to 15-m-d 1 .

From low flows measured in all seepagemeters, the estimated lagoon bottom sedimentmean conductivity was estimated as 0.015 m-d 1 , at least an order of magnitude lower thanriparian soil K s reported above. This impliesthat probably most of the flow to and from theocean ( QGS ) would occur through riparian

zones, if there was a head difference betweensea and lagoon water levels.

3.2. Lagoon water level fluctuations

During the disconnected phase of thesummerautumn of 2006 (November to June),lagoon levels varied almost negligibly, as seen

in Fig. 3 (NB: sea level was converted into rela-tive height above our reference point, whereaverage sea level was 98.5 m). The only excep-tions were on days of high winds (Fig. 3a). Inearly winter, due to rainfall events, Yali stream-flow eroded the bar that had closed the streammouth during the previous dry season (Fig. 3b).Then, the lagoon and the ocean were connected through the outlet, inducing a daily oscillationin the lagoon water level (Fig. 3b, c). Disconnec-tion during the study period occurred during mid-September. About a week before closure, the

periodic fluctuation of water levels was replaced by stepwise increases, until water levels stabilized circa 18 September with the bar finally closed

(Fig. 3d).Conversely, during the connected phase thelagoon-level amplitude difference averaged 20 cm, varying between 8 and 80 cm ( + 2 cm),depending on the tide. An analysis of the aver-age pattern (Fig. 4) for 10-min data from 10June to 15 September showed that it typicallytook 4 h to reach peak lagoon level after asea-level threshold was surpassed by the risingtide (related to the outlet base level, as discussed further below). This peak was followed by sealevel dropping below lagoon level, just after high tide, in the sea recession limb (residualerrors in the average pattern with mean 0.0137and standard deviation 0.0705). For the reces-sion that followed, it took approximately 13 hto return to initial levels, when a new inundationcycle was initiated.

Spectral analysis of lagoon (Fig.5) vs. sealevels showed that the most important frequen-cies in decreasing order are 12.5 h (high tides),

24.2h (tidal daily cycle) and 16.4 days (mooncycles). Sea data revealed a peak at 12.5h and another lesser one at 23.7 h (not shown).

When the outlet bar formed and the lagoondisconnected from the ocean, there was no appa-rent relationship between sea level and lagoonlevel; for example, spectral analysis only revealsa slight peak at period 0.997 days. The lagoon



7/13


8/13

level varied little during the day (only a fewmillimetres) for any tide amplitude, as exempli-fied for February 2006 (Fig. 3d).

However, an exception occurred on 13February, attributed to strong winds which prob-

ably produced a slope in the lagoon level. Thesewind effects were most pronounced during thedisconnected phase. An analysis of wind datashowed that winds were stronger when wind direction was from the SE, closer to the lagoonlongitudinal axis, and directed towards thetransect area (correlations not shown, average

0.715), as well as the reverse direction, when adecrease in the level occurred, attributed to anopposite sloping effect. Correlations were nega-tive when wind direction was perpendicular tothe lagoon, indicating that the slope went

upstream instead. Further studies are needed,including monitoring with a transducer in theextreme SE of the lagoon.

Based on the data gathered, rainfall had littleeffect over lagoon water level, except for thoseevents that opened the outlet bar in the dis-connected situation. In sum, the overriding

98.60

98.65

98.70

98.75

98.80

98.85

98.90

98.95

00:00 03:00 06:00 09:00 12:00 15:00 18:00 21:00 00:00 03:00

Time (hours)

R e

l a t i v e

h e

i g h t ( m )

Fig. 4. Typical behaviour pattern in lagoon water level periodical oscillations based on 10 min data from June 10 toSeptember 15, 2006.

f (cycles/time unit)

S p e c

t r u m

0 0.1 0.2 0.3 0.4 0.50

10

20

30

40

50

60

Fig. 5. Spectral analysis for lagoon water level from July 13 to August 31, 2006, based on hourly data and time unit of 24 h.



9/13

influence on lagoon level was sea-level fluctua-tions. Nevertheless, with rainfall the groundwater level between lagoon and ocean increased veryquickly (Fig. 3c), probably due to the highly con-ductive porous media.

3.3. Water level numerical analysis

A Fourier analysis performed in Matlab TM

(Fig. 6) applied to hourly sea and lagoon timeseries revealed that a simple impulse responsefunction of exponential form exp( t /2.5) fitthe sealagoon relation very well, with an appro-

priate threshold related to the outlet base level(98.67m, i.e. approximately 1520 cm above

mean sea level during connected phase). Thisis supported by our field physical interpretationof the level data series (Fig.3): when the lagoonlevel reached a peak, it corresponded to the max-imum sea level, and initial recession was con-trolled by fast outflow through the outlet withdescending sea level. But once the lagoon levelreached outlet level (and if sea level was lower than outlet base level, at 98.67 m in Fig. 3), therecession took place at a slower rate, probablylinked to slower porous media flow throughthe sand berm to the ocean (Fig.3). Another pos-sibility is that once the level drops below a cer-tain level, stream water flows in, dampening therecession drop; this would be improbable during

the connected phase due to negligible dry seasonstreamflow.

Modelling simulations for water levels in thecentral transect, using the Dupuit approximationgiven in Eq. 1, were done for the range of k val-ues measured.

To include partially some adjustment due todensity dependence, water levels for the boun-dary conditions (sea and lagoon level data)were corrected as a linear function of salinitycontent, using the GhybenHerzberg equationthat provides an empirical first approximationin natural (more or less static) coastal aquifers[23]. This approach resulted in multipliers1.0291 and 1.0270 for sea and lagoon hydraulic

heads, respectively. Initial conditions were aninterpolation between both boundary conditionson July 14, 2006, which explains why it took 23days for the initial condition error to propagateand finally diffuse, as seen in Fig. 7.

Figure 7 shows results for the best-fit param-eter values in the high-end range for k (app-roximately 10 4 m s 1 , in accordance with thevalues reported; see Section 3.1). The resultsadjusted well for the general trends in amplitudeoscillations and periods. The smaller scalemismatch is attributed to differing salinities

between boundary conditions and possiblyother density-dependent effects. Additionally,the vertical beach boundary condition might be

98.6

98.7

98.8

98.999

99.1

99.2

99.3

99.4

14-07 16-07 18-07 20-07 22-07 24-07 26-07 28-07 30-07

Time (hours)

R e

l a t i v e

h e

i g h t ( m )

Simulated

Real

Fig. 6. Fourier analysis for lagoon water level from July 13 to August 31, 2006, based on hourly data.



10/13

affecting this kind of simulation. Given the aqui-fer estimated depth of 100 m and soil porosity of 0.43, the diffusivity coefficient D, computed from Eq. 3, was 2.3 10 2 m2 /s.

3.4. Water quality

Concerning water quality, groundwater pHaveraged 7.1 (range: 6.5 to 7.6), while lagoon and

sea values were very stable, 8.6 and 8.1, respec-tively. Dissolved oxygen did not show a defined spatial pattern (groundwater 24 mg L 1 , sea and lagoon 79 mgL 1 , with a maximum in the latter of 12mg L 1 ).

Groundwater discharge from upslope into thelagoon was fresh (average conductivity lower than 0.7mS cm 1 , increasing to 45 close to the

lagoon). Annual mean conductivity for the lagoonwas 51.9mS cm 1 for 2006, with values rangingfrom20to100mScm 1 (highervalues in summer as would be expected). Increased salinity/conduc-tivity in the lagoon was linked to seawater intru-sion and evaporation in the shallow system,known to happen in this type of system [28].

Phosphate in groundwater exhibited (data notshown) a gradient of progressive decrease

towards the sea with some very high valuesupslope from the lagoon and in nearby drainageditches discharging into the lagoon, over 1.5mg/L P-PO 4 . Possible sources include agriculturalfertilizers, organic deposits of past poultry oper-ations, as well as cattle illegally introduced into the wetland area for grazing. Lagoon phos-

phate concentrations, however, showed a narrow

a)

100.8

101.0

101.2

101.4101.6

101.8

102.0

102.2

102.4

12-07 17-07 22-07 27-07 01-08 06-08 11-08 16-08 21-08 26-08 31-08

R e

l a t i v e

h e

i g h t ( m )

101.0

101.2

101.4

101.6

101.8

102.0

102.2

102.4

12-07 17-07 22-07 27-07 01-08 06-08 11-08 16-08 21-08 26-08 31-08

R e

l a t i v e h e

i g h t ( m )

Model (Ks 6.2 10 5)

CP2

b)

Model (Ks 1.7 10 4)

CP5

Fig. 7. Modelling results (thin line; saturated hydraulic conductivity K s reported) vs. piezometer data (thick line) com- parison of groundwater levels for the sand bar between the lagoon and the sea, for the central transect piezometers (seeFig. 2): (a) CP2 (closer to lagoon in riparian vegetation area; note that since 101.7 is the maximum recordable relativeheight by this piezometer, there is an artificial plateau in the graph); (b) CP5 (close to topographic peak of sand dune,nearer the seaside).

211 A. Dussaillant et al. / Desalination 246 (2009) 202214


11/13

range: 0.10.5mg/L P-PO 4 . Nitrate results areunder revision due to possible interferences inthe method, and are not reported here; initial evi-dence suggests that this is the limiting nutrient inthe lagoon and riparian area [18].

4. Conclusions

Based on our initial observations and datacollection, Yali lagoon water-level dynamicsappear to be driven by sea level and streamflowvariability. Our results suggest that thesefactors can be organized hierarchically. Theupper hierarchical level is seasonal: lagoonlevel responds to the lagoon surface water

being connected to or disconnected from theocean through the outlet.During disconnection (summer-autumn), the

lagoon level is generally independent of sealevel, even though there is a 150 m poroussand bar between the two. Apparently, dampen-ing is enough to minimize the effect of ground-water pressure wave. The bar is occasionallyovertopped by storms, which could affect lagoonwater level temporarily. Nevertheless, suchevents were not observed during our data collec-tion in 2006.

Conversely, during the connection phase twolower level temporal scale hierarchies are rele-vant: monthly (moon cycles every 16.4 dayson average) and daily (tidal cycles every 12.5and 24.2h), as the pattern, spectral and Fourier analyses have shown.

Wind effects can be important particularlyduring the disconnected phase, while in the con-nected phase, these are overridden by the ocean-

level factor. Rainfall events seem to be not animportant driver for lagoon level, except for those rains that change the phase from discon-nected to connected, that is, that interact withthe higher temporal scale hierarchy.

Numerical analysis results show that a simpleexponential impulse response function relatesthe sea forcing to lagoon level during the

connected phase. Additionally, a simple Dupuitnumerical model gives a reasonable approxima-tion of piezometer water-level trends betweenthe sea and the lagoon, even though there are sal-inity and beach slope effects that deserve further attention. Also, as we have observed since 2003,sea storms, El Nin o events and general interan-nual variability will affect water exchanges, sowe plan to continue monitoring and incorporat-ing new data into our analyses.

Water quality monitoring has shown the potential presence of a double salt wedge, aswell as a phosphate high-concentration zone

just upslope of the lagoon, that likely relates toillegal cattle entrance and past poultry operation

waste. This will be investigated further since it isa cause of concern for the management of thisunique coastal wetland.

In the near future, we hope to use the tools pre-sented here, complemented with additional ones(e.g. complete water balance, density-dependentmodelling, extensive data gathering) and findings

by an ecology team working in the area, to con-tinue our study of this system, examining topicssuch as water level versus plant distribution rela-tionships for different scenarios in sea- and lagoon-level dynamics.

Acknowledgements

Authors would like to thank our collaborator,ecologist J.M. Farin a, and the managers of CONAF. Data were provided by the Navy,Direccion Meteorologica de Chile and DireccionGeneral de Aguas. M. Sepu lveda and Agrosuper S.A. (through C. Vives) provided private reports.

Useful suggestions for data analysis were pro-vided by B. Fernandez, W. Palma, P. Irarrazavaland R. Cienfuegos. Special thanks are due toP. Pasten for providing the equipment. Wewould like to thank also A. Packman, E. Barthe-lemy and S. Tyler for fruitful field discussions,as well as J. Aravena, F. Varas and E. Mignotand all the people who helped us in the coastal

A. Dussaillant et al. / Desalination 246 (2009) 202 214212


12/13


13/13

P. Universidad Catolica, Chile (preliminary draftin Spanish), 2008.

[25] J.D. Salas, J.W. Delleur, V. Yevjevich and W.L.Lane, Applied Modeling of Hydrologic Time Series ,Water Resources Publications, Littleton (CO),USA, 1980.

[26] W.J. Weber and F.A. Digiano, Process Dynamicsin Environmental Systems , Wiley, New York,1996, 943 pp.

[27] MOP Informe tecnico N 8 421: Area de restriccionsector hidrogeologico de aprovechamiento comu nde Yali bajo El Prado. Ministerio de Obras Pu bli-cas, Gobierno de Chile, Government Report, 2005.

[28] E. Diamantopoulou, M. Dassenakis, A. Kastritis,V. Tomara, V. Paraskevopoulou and S. Poulos.Seasonal fluctuations of nutrients in a hypersalineMediterranean lagoon, Desalination , 224(2008),271279.

A. Dussaillant et al. / Desalination 24 (2009) 20 21214 6 2 4

Date post:	02-Apr-2018
Category:	Documents
Upload:	andre-porfirio1
View:	216 times
Download:	0 times

Dussaillant, 2007

Documents