ContentslistsavailableatScienceDirect
Journal
of
Computational
Science
j o u r n al ho me p a g e :w w w . e l s e v i e r . c o m / l o c a t e / j o c s
The
impact
of
Sea
level
rise
in
the
guadiana
estuary
Lara
Mills
a,∗,
João
Janeiro
a,
Antonio
Augusto
Sepp
Neves
a,
Flávio
Martins
a,b aCentrodeInvestigac¸ãoMarinhaeAmbiental(CIMA),UniversityofAlgarve,Faro,PortugalbInstitutoSuperiordeEngenharia(ISE),UniversityofAlgarve,Faro,Portugal
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:Received8January2020
Receivedinrevisedform8May2020 Accepted8June2020
Availableonline12March2020 Keywords:
Guadianaestuary Sealevelrise MOHID Salinityintrusion
a
b
s
t
r
a
c
t
Understandingtheimpactofsealevelriseoncoastalareasiscrucialasalargepercentageofthepopulation liveonthecoast.Thisstudyusescomputationaltoolstoexaminehowtwomajorconsequencesofsea levelrise:saltintrusionandanincreaseinwatervolumeaffectthehydrodynamicsandfloodingareasof amajorestuaryintheIberianPeninsula.A2DnumericalmodelcreatedwiththesoftwareMOHIDwas usedtosimulatetheGuadianaEstuaryindifferentscenariosofsealevelrisecombinedwithdifferent freshwaterflowratesconsideringvaryingtidalamplitudes.Anincreaseinsalinitywasfoundinresponse toanincreaseinmeansealevelinbothhighandlowfreshwaterflowratesatallareasaroundtheestuary. Anincreaseinfloodingareasaroundtheestuarywasalsopositivelycorrelatedwithanincreaseinmean sealevel.
©2020TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).
1. Introduction
Ariseinmeansealevelisaglobalconcernas10%oftheworld’s
populationlivewithin10melevationofthecurrentsealevel[3].
BasedonprojectionsbytheIntergovernmentalPanelonClimate
Change(IPCC),therateofsealevelriseisaccelerating.An
accel-erationinglobalsealevelrisehasbeenreportedthroughoutthe
20thcenturybasedonsatellitealtimeterdata[6].Globalmeansea
levelrisecouldbeashighas1mbytheyear2100ifgreenhouse
gasemissionscontinuetobeveryhigh[5].Asthenumberof
peo-pleaccommodatingcoastalareasincreases,anaccelerationinthe
rateofsealevelrisewillseverelyimpactsocietyandtheeconomy.
Impactsofsealevelriseoncoastalareasincludesubmergenceof
land,increasedflooding,increasederosion,changesinecosystems
andanincreaseinsalinity[16].Notyetfullyexploredinthe
litera-tureishowthedynamicsofsalinityinestuarieswillevolvedueto
sealevelrise.Itisthustheaimofthisworktoquantifya
relation-shipbetweensealevelriseandsalinitycontentinestuaries,asan
increaseinestuarinesalinityhasthepotentialtodamageestuarine
environments.
Fromaphysical standpoint,anestuaryis definedasa
semi-enclosedbodyofwaterwhererivermeetssea,extendingintothe
riverasfarastheupperlimitofthetides.Estuariesare
produc-tiveenvironments,hostinghighlydiverseandcomplexecosystems
∗ Correspondingauthor.
E-mailaddresses:a60162@ualg.pt(L.Mills),jmjaneiro@ualg.pt(J.Janeiro), aaneves@ualg.pt(A.A.S.Neves),fmartins@ualg.pt(F.Martins).
[20].Fisheries,aquaculture,ecotourism,andportfacilitiesrelyon
estuariestocontributerevenuestonational, regional,and local
economies.
Thesetransitionalareaswherefreshwaterderivedfromland
mixeswithsaltwaterfromtheseaaresensitivetochangesin
cli-mate suchas sealevel rise.Estuariesadapt tosealevelriseby
deepeningtheirchannels.Thisincreasestheaccommodationspace
ofsediments,furtherenhancingebbandfloodasymmetry[20].
Awatercolumnisconsideredstratifiedwhenlessdensewater
liesoverdenserwater.Thus,thestratificationofanestuaryis
deter-minedbythebalance betweenbuoyancyand mixingof denser
saltwaterfromtheseaandlessdensefreshwaterfromland.An
increaseinsealevelallowsmoresaltwatertoentertheestuary
[19].Thus,amajorconsequenceofsealevelriseinestuariesisthe
intrusionofsaltfromseawaterintofreshgroundwater,whichcan
furtherimpactstratification[15].
Theintrusionofsaltwillhaveaprofoundeffectonthephysical
propertiesofestuariesasthedynamicmixingofsaltwaterandfresh
wateristhedrivingfactorregulatingstratification[19].Changesin
oceanicsalinitycombinedwithvaryingflowratesoffreshwater
dischargeintotheestuarywillaltertheentireestuarinecirculation
[19].HongandShen[11]foundanincreaseinstratificationinthe
ChesapeakeBayasaresultofanincreaseinsalinitybasedonthe
resultsofathree-dimensionalnumericalmodel.Thischangein
hor-izontalsalinitygradientswillfurtheralterestuarinecirculationand
causeoxygendepletion[11].Alterationsinestuarinecirculation
willbedetrimentaltomarineecosystemsthatcannottoleratehigh
salinitycontent[4].Furthermore,sealevelriseincreasessalinity
upstreamandimpactstidalcurrents[11].Astudyontheimpactof
https://doi.org/10.1016/j.jocs.2020.101169
2 L.Mills,J.Janeiro,A.A.S.Nevesetal./JournalofComputationalScience44(2020)101169
anincreaseinsalinityonLouisianaestuariesbyWisemanetal.[22]
foundbothpositiveandnegativetrendsattributedtothe
variabil-ityoffreshwaterflowrates.Theseauthorsfoundthatanincrease
insalinityresultsin thedeathor declinein theproductivityof
marshes,whicheventuallyleadstolandloss[22].Furtherimpacts
ofsaltintrusionincludecontaminationofwatersuppliesforboth
consumptionandfortheindustry[10].
Theaimofthisarticleistopresenttheresultsofanumerical
modelsimulatingsaltintrusionintheGuadianaEstuary,amajor
estuaryontheIberianPeninsula. Themodelsimulatesdifferent
sealevelrisescenariosasprojected bytheIPCCalongwith
dif-ferentriverdischargeflowratesandvarioustidalamplitudes.As
changesinsealevelinevitablyalterthetopographyofestuaries
[17],thesimulationsofthepresentstudywererunovertwo
pos-siblefuturebathymetriesoftheGuadianaEstuarytodetermineif
actionmustbetakentomaintainthecoastlineortoallowflooding
ofsurroundingareasoftheestuary.
2. Methods
2.1. Studyarea
TheGuadianaEstuaryconnectstheGuadianaRivertotheGulfof
Cadiz.TheGuadianaRiveristhefourthlongestriverontheIberian
Peninsulawithalengthof810km,200ofwhichforma border
betweenPortugalandSpain[7].Beginningatitsmouthinfront
ofVilaRealdeSantoAntóniointheAlgarveregionofPortugaland
Ayamonte,Spain,theestuaryextendsabout80kmnorthtoitstidal
limitatMértola.Theestuarydrainsatotalareaof66,960km2[9]
infrontofthesehighlypopulatedregions.TheGuadianaEstuaryis
atitswidestinfrontofVilaRealdeSantoAntóniowithawidthof
800mandismostnarrownearMértolawithawidthof70m.The
averagedepthoftheestuaryis5m,butinsomecases,itcanreach
10m.
Downstream,theestuaryisintheformofasubmergeddelta
wherethereismoderatewaveenergy [9]. Theestuaryis
classi-fiedasarock-boundestuaryduetoitsnarrowandrelativelydeep
channel,locatedonapassivemarginwherethevolumeofwater
enteringtheestuaryduringthefloodtideislargerthanthe
freshwa-terdischargefromtheGuadianaRiver[9].TheGuadianaEstuaryis
characterizedbyasemi-diurnalmeso-tidalregimewithtidal
influ-ence80kmupstreamfromwhereitmeetstheGulfofCadiz[7].The
averageneaptidalrangeis1.28mandtheaveragespringtidalrange
is2.56mwithamaximumspringtidalrangeupto3.44m[9].
Tidesandfreshwaterinputcontroltheverticalmixingand
strat-ificationof theestuary, which rangesfrom very well-mixedto
partiallystratified[9].Themixingoftheestuaryvariesbyseason,
withawell-mixedwatercolumnwhenthereislowdischargefrom
theGuadianaRiverduetoalackofrainfallandapartiallystratified
watercolumnwhenthereismoredischargefromtheriver.Itis
onlyunderextremeconditionsthatthewatercolumncanbecome
stratified[9].Theebbcurrentsoftheestuaryaregenerallyfaster
thanfloodcurrents,eveninlow riverconditions,aneffectmost
likelyduetothelargehydraulicdepthoftheestuarinechannel.The
ebb-dominancebecomesmorepronouncedastheriverdischarge
increases[9].
The freshwater discharge of the Guadiana River can range
between10m3/sand4660m3/s.Theconstructionofover100dams
sincethe1950shascontrolled70%ofthedrainagebasinbytheend
of2000[9].OfrelevanceistheAlquevaDam,located60kmfrom
theheadoftheestuary.Itisthelargestreservoirinsouthernand
westernEuropestoring4150hm3ofwater[9].Itsclosurein2002
hassignificantlyreducedthemeanriverflowfromanaverageof
143m3/soverthelast26yearsto16m3/safter2002[9].
Down-streamthedam,theriverhasephemeralflowsandisdryfor40%
oftheyear[8].TheGuadianaEstuarycurrentlyfacesareduction
inriverflowduetoscarcerainfallandanincreaseinfreshwater
storage.AmapofthestudyareaisshowninFig.1.
2.2. Hydrodynamicmodelsetupandsimulatedscenarios
The MOHID water modelling system was used to simulate
the hydrodynamicsof theGuadiana Estuary and the effects of
differentsealevelrisescenariosonsalinitydistributionand
trans-port. MOHIDis a three-dimensional hydrodynamical modelling
system that integrates diversenumerical models in an
object-orientedprogrammingapproach[13].MOHIDsimulatesphysical
andbiogeochemicalprocessesinthewatercolumnandsediments
[13]. MOHID hasbeen used to simulatethe hydrodynamics of
manycoastalandestuarinesystems,mainlythemajorestuariesof
Portugal,butalsointheNetherlands,FranceandIreland[18].The
modelhasbeenpreviouslyvalidatedandcalibratedintheGuadiana
Estuarybycomparingin-situmeasurementsofsalinitywithmodel
results[12].ArecentstudybyQuesadaetal.[18]usedMOHIDin
2Dbarotropicmodetosimulatetheeffectsoffreshwaterdischarge
andtidalforcingontidalpropagationintheGuadianaEstuary.This
modelaccuratelyproducedphasevaluesoftidalelevationsand
cur-rentsasconfirmedbyaharmonicanalysis[18].Themodelwasable
todemonstratehowfreshwaterdischargehasastrongereffecton
tidalwaveamplitudeupstreamtheestuary,whereastidalforcing
controlstidalwaveamplitudedownstreamtheestuary[18].
ThemodelsolvestheNavier-Stokesequationsconsideringthe
hydrostaticandBoussinesqapproximations.uiisthevelocity
com-ponentineachdirection;fistheCoriolisparameter,=0+’isthe
densitydecomposedinitsreferencevalueplusanomaly,andAiis
theturbulenceviscosityineachdirection.Theequationsare
dis-cretizedusingthefinitevolumemethod.Thesolutionschemeis
semi-implicitoftheAlternatingDirectionImplicit(ADI)type.The
S21schemeofAbbottetal.[1]forshallowwaterequationswasused
andadaptedto3DaccordingtoMartinsetal.[14].Theequations
areasfollows:
∂
u1∂
x1 +∂
u2∂
x2 +∂
u3∂
x3 =0∂
u1∂
t +∂
uju1∂
xj =fu2−gn 0∂
∂
x1 − 1 0∂
ps∂
x1 − g 0 z∂
∂
x1 dx3 +∂
∂
xj Aj∂
u1∂
xj∂
u2∂
t +∂
uju2∂
xj =−fu1− gn 0∂
∂
x2 − 1 0∂
ps∂
x2− g 0 z∂
∂
x2 dx3 +∂
∂
xj Aj∂
u2∂
xj∂
p∂
x3 = −gThesamediscretizationandcomputationalsolutionwereused
totransportsalinity,whichisassumedtobeapassivetracer.
Thecalculationtimetostabilizethesalinityfieldisveryhigh
whentheriverflowislowduetothehighresidencetimeofthe
estuaryinlowriverdischargeconditions.Forexample,withariver
flowof10m3/sittakestwomonthsofsimulatedtimeforthe
salin-ityfieldtostabilize.Thecomputationalmeshwaschosentoprovide
a spatialresolutionappropriate forthestudywithoutincurring
excessivecalculationtime.Thus,eachsimulationusedaCartesian
Fig.1. Mapof(a)theIberianPeninsulastudyareaand(b)acloserlookattheGuadianaEstuary.Orangedotsin(b)markwheresalinity,velocityandelevationtimeseries werecollected.Bluedotsmarkreferencepointsinthestudyarea(Forinterpretationofthereferencestocolourinthisfigurelegend,thereaderisreferredtothewebversion ofthisarticle).
of30mtoevaluatesalinitydistributionandtransportaswellas
coastalfloodingatthemouthoftheGuadianaEstuary.The
simula-tionswereruninanIntelXeonGold6138serverandthecodewas
parallelizedusingOpenMPanddistributedover8cores.Withthis
setup,twomonthsofsimulatedtimetakesmorethanfifteendays.
The model incorporates various sea level rise scenarios
combined with different river discharge scenarios typical of
the Guadiana River over two separate bathymetries. The first
bathymetrywascomputedbytriangularinterpolationofmeasured
bathymetricdataonthemeshof1400×350cellswithaspacestep
of30m.Thisbathymetryisindicativeofthepresentbathymetryof
theestuaryandcorrespondstoacaseinwhichmaximumhuman
interventionistakentokeepthecoastlineunchanged.Thesecond
bathymetryrepresentsasituationinwhichthereisnohuman
inter-vention,thusallowingmaximumfloodingofthesaltmarshesand
low-lyingareas.Process-basedmodelssuchasMOHIDcannot
eval-uatelong-termgeomorphologicaldynamicssotheresultsfroma
behavior-orientedmodelbySampathetal.[21]wereusedforthis
specificbathymetry.Thebathymetrywascomputedbasedonthe
rateofsealevelrise,accommodationspaceforsedimentdeposition
andverticalaccretionofsedimentsduetoinundation[21].Thetwo
differentbathymetriesareshowninFig.2.
Twodifferentfreshwaterdischargeswereimposedinthemodel
representing:(1)alowfreshwaterdischargeof10m3/swhenthere
islittlerainand(2)ahighfreshwaterdischargeof100m3/swhen
thereismorerain.Themodelfurtherexploresthreevaryingtidal
amplitudesoftypeM2representing:(1)typicalsystemconditions
equivalenttotheaveragetideconditions,resultinginanamplitude
of1m;(2)springtideequivalenttoanamplitudeof1.28mand
(3)neaptideequivalenttoanamplitudeof0.64m. Tosimulate
thesalinitytransportintheGuadianasystem,thesalinityofwater
fromtheGuadianaRiverwasconsiderednullandthesalinityofthe
coastalregionadjacenttotheestuarywassettoaconstantvalue
of36.Thesevaluesaretypicalforthisregion.Therearenounitsfor
salinity,inagreementwithrecommendationsbyUNESCO.
Thetwodifferentriverdischargescenarios mentionedabove
were combined with different sea level rise scenarios derived
fromsealevelrise projectionsprovided bythe5th IPCCreport
forRepresentativeConcentrationPathway(RCP)4.5andRCP8.5.
RepresentativeConcentrationPathwayreferstogreenhousegas
concentration[5].The5threportoftheIPCCprojectedseveral
pos-siblevaluesinglobalmeansealevelriseforthesetwoRCPvalues
overtheyears2046–2065,2081–2100and2100.Duetothehigh
computationalcostandthefactthatsealevelrisescenarioswould
becombinedwithtwodifferentriverdischargescenariosaswell
asthreedifferenttidalamplitudes,itwasnotpossibletosimulate
allsealevelriseforecastsfromtheIPCC.Itwasdecidedtoconsider
threecasesofsealevelrisethatcoveredallpossibilitieswithinall
scenariosfromtheIPCC.Thesealevelriseprojectionscorrespond
toasealevelriseof0.24m,0.48m,and0.79mcorrespondingto
theyearsof2040,2070and2100,respectively.Thevalueswere
imposedinthemodelbychangingthereferencelevelofthetidal
components.
Threedifferentsealevelrisescenariosalongwiththecurrent
sealevelwerecombinedwiththetworiverdischargescenarios
along twopossiblebathymetriesoftheestuary. Additionally,in
ordertoassessthefuturestateoftheGuadianaEstuarywithrespect
tospecifictidalevents,springandneaptidescenarioswerealso
simulated.Theseboundaryconditionsonthemodelproducean
evaluationofthepreviouslymentionedscenariosresultingin
pro-jections,notforecastsofthefuturestateoftheGuadianaEstuary.
Whileitisexpectedforthetidesandriverfluxestochangedueto
sealevelrise,theaimofthisstudyistoanalyzethetrendsproduced
bysealevelriseonspringandneaptides.Therefore,theresults
provideanevaluationofthesensibilityoftheGuadianaEstuaryto
changesintidalamplitudecombinedwithsealevelrise.
Each sea level rise, river discharge and tidal scenario were
imposedinthemodeltoproduceseveraltimeseriesandhorizontal
distributionmaps.Thelocationsforeachtimeseriescanbeseen
inFig.1,correspondingtotheintersectionoftheGuadiana
Estu-aryandEsteirodaLeziria(381147),inthemiddleofEsteiroda
Leziria(39196)andatthewesternmostregionofEsteiroDaLeziria
(39223).
Thetimeserieswereconstructedoveraperiodof2tidalcycles
(24.83h)andrepresentthetemporalevolutionofaveragevalues
4 L.Mills,J.Janeiro,A.A.S.Nevesetal./JournalofComputationalScience44(2020)101169
Fig.2.ComputationalbathymetryoftheGuadianaEstuaryforamaintainedcoastline(left)andanunmaintainedcoastlinethatallowsfloodingofthesaltmarshesand low-lyingareas(right).
Table1
Classesoffloodingtimeduringonetidalcycle.
Class FloodIntervalInHours
1 0–1.2 2 1.2–2.4 3 2.4–3.7 4 3.7–4.9 5 4.9–6.1 6 6.1–7.3 7 7.3–8.6 8 8.6–9.8 9 9.8–11.0 10 11.0–12.2
respectiveminimum/maximumvaluesobservedwithinthe
sim-ulatedscenarios.Average,maximum,andminimumvalueswere
obtainedbygroupingallthree tidalamplitudes combinedwith
bothfreshwaterflowratestoexaminethedifferencesbetweenthe
presentand2100.Scenariosbasedonthepresentbathymetrywere
analyzedseparatelyfromthosesimulatedontheflood-permitting
bathymetry.
Floodingareaswerecomputedfortendifferentclassesof
sub-mersiontimeduringonetidalcycleoverthebathymetryallowing
maximumflooding.Table1outlinesthedifferentfloodingclasses.
ForeachclassinTable1theareaoffloodingwascomputed,and
ahistogramwasconstructedforthevariousscenarios.Flood
distri-butionmapswerecreatedtoshowtheamountoflandsubmerged
duringatidalcycleaccordingtofloodingclass.
3. Results
Theresultsofthemodelaredisplayedasfollows:section3.1
includestheresultsofthreetimeserieslocationsonthewestern
marginoftheestuaryinthesaltmarshesofCastroMarim,section
3.2includesthehorizontaldistributionofsalinityaswellaschanges
insalinityintrusionlengthinresponsetosealevelrise,section3.3
exploressalinitydistributionatspringandneaptide,andsection
3.4examinesfloodingareasduetosealevelrise.
3.1. Timeseriesofvelocity,salinityandelevation
Figs. 3, 4 and 5 show the time series results at the
previ-ouslymentioned locationsfor boththepresent bathymetryand
bathymetrythatallowsflooding.
The time series results for the three locations downstream
EsteirodaLeziria(381147,39196and 39223),ascanbeseen
in Figs. 3–5, portrayan increase insalinity from 2019to2100
duetosealevelrise.Theincreaseinsalinityisdirectlycorrelated
withwaterelevation:salinityvaluesarehigherwhenwaterlevelis
higherandsalinityvaluesarelowerwhenwaterlevelislower.The
locationclosesttothemainchannel(381147)containsthehighest
salinityvaluesandmostvariabilityoverthe2tidalcycleperiodas
canbeseenin
Fig.3.
Locationsfurtherwest(39196and39223)containsmaller
val-uesinsalinityandlessvariabilityduringthe2tidalcycleperiod.
Theincreaseinsalinityforalllocationsismorepronouncedforthe
bathymetrythatallowsflooding,withsalinityincreasingbyabout
4witha0.79mriseinsealevel.Velocitydoesnotchange
signifi-cantlyfromthepresentto2100butFigs.3–5showaslightdecrease
invelocityduetothe0.79msealevelrise.
3.2. Horizontaldistributionofsalinitytransport
Thesalinitydistributionmapsforthepresent bathymetryin
highwaterconditionsforthevariousscenariosofsealevelrisecan
beseeninFig.6andforlowwaterconditionin
Fig.7.Onlysalinitymapsforariverdischargeof10m3/sand
100m3/salongwithaverageM2tidalamplitudeof1mareshown
astheyaremostrepresentativeoftypicalsystemconditions.
Forscenarios withalow freshwaterdischarge (10m3/s)and
athightide,thesalinityprogressesupstreaminresponsetosea
inter-Fig.3. Timeseriesresultsofsalinity,velocity,andelevationat381147forthepresentbathymetry(left)andbathymetryallowingflooding(right).
Fig.4. Timeseriesresultsofsalinity,velocity,andelevationat39196forthepresentbathymetry(left)andbathymetryallowingflooding(right).
sectionofthemainchannelandCarresqueiraandLezíriawhere
thesalinityhasa valueof20, salinityprogresses about3100m
upstreamfromthepresentsituationtotheyear2040.The
salin-ityintrusionaftertheyear2040continuestoprogressupstream,
butatamuchslowerrate.Theevolutionofsalinityatthemouthof
theestuaryislessdramaticasmentionedinthetimeseries
analy-sis,butsalinityintrusioncanevolvebytensorhundredsofmeters
Fig.8.
Thedistanceofsalinityintrusionatspringand neaptidefor
thetwo freshwaterflowrateswasalsocomputedforboth high
waterandlowwaterconditions.Table2showstheincreaseinsalt
intrusionlengthfromnowto2100forbothtypesofbathymetry.
Allcases of tide, freshwater flow, and bathymetry result in
an increase in salt intrusion length due to sea level rise. The
increase is much more significant for the bathymetry allowing
flooding compared to the increase in salinity for the present
bathymetry. The increase in salinity intrusion length is also
higherfor thelowfreshwater discharge.Thefurthestextension
of salinity intrusion is at spring tide with a freshwater flow
of 10m3/s where the salinity intrusion length increases from
26.31km to28.14km upstreamthe mainchannel fromnow to
2100.
3.3. Salinitydistributionatspringandneaptide
Salinityclassesforspringandneaptidescenarioswere
com-putedbasedontheamountofareacontainingsalinityvalueswithin
theintervalsasshowninTable3.
Theresultsdemonstrateanoverallincreaseinareacoveredby
highersalinityvalueswithrespecttosealevelriseforallcasesof
salin-6 L.Mills,J.Janeiro,A.A.S.Nevesetal./JournalofComputationalScience44(2020)101169
Fig.5.Timeseriesresultsofsalinity,velocity,andelevationat39223forthepresentbathymetry(left)andbathymetryallowingflooding(right).
Fig.7. Salinitydistributionmapsinlowwaterconditionsforadischargeflowof10m3/s(left)and100m3/s(right)forthepresentyear,2040,2070and2100.
Table2
Differencesinsalinityintrusionlengthfromthepresentto2100.
SalinityIntrusionLengthIncreasefromPresentto2100(km)
HighTide LowTide
Present Flooding Present Flooding
SpringTide10m3/s 0.99 1.83 0.52 2.10
SpringTide100m3/s 0.26 1.23 0.13 1.70
NeapTide10m3/s 1.55 1.73 0.79 2.10
NeapTide100m3/s 0.12 2.00 0.47 2.29
Table3
Changesinsalinityfromthepresentto2100forspringandneaptide. AreasbySalinityClassfrom
Presentto2100(km2)
SalinityClass Spring100m3/sHighTide Spring
100m3/s LowTide Spring 10m3/s HighTide Spring 10m3/s LowTide Neap 100m3/s HighTide Neap 100m3/s LowTide Neap 10m3/s HighTide Neap 10m3/s LowTide 0−5 −3.49E+00 −4.09E+00 −1.09E+00 −1.22E+00 1.13E+00 2.76E+00 −9.52E-01 −9.68E-01 5-10 2.36E+00 3.56E+00 −1.08E-02 −6.84E-02 −2.42E+00 −2.35E+00 −3.27E-01 −8.22E-01 10-15 9.93E-01 2.66E+00 −2.45E-01 −5.25E-01 4.47E+00 4.15E+00 4.06E-01 1.31E+00 15-20 −3.49E-02 6.36E-01 −5.00E-01 1.13E-01 2.20E+00 2.55E+00 −4.37E+00 −5.28E+00 20-25 −4.45E+00 −4.06E+00 −5.56E-02 −2.47E-01 7.20E-02 1.07E-01 3.17E+00 4.86E+00 25-30 5.14E+00 6.96E+00 −6.81E+00 −8.59E+00 1.85E+00 7.56E-02 6.79E+00 6.97E+00 30-35 5.64E+00 4.87E-01 9.16E+00 1.58E+01 8.45E-02 8.45E-02 2.58E+00 1.23E+00 35-36 2.00E-01 2.00E-01 5.91E+00 1.08E+00 0.00E+00 0.00E+00 8.27E-02 8.27E-02
ityclassbetweenthepresentand2100.Thedifferencesaremore
pronouncedathightideandforlowerfreshwaterflows.
3.4. Areasofflooding
Floodingareawascomputedasa functionof thenumber of
hoursofsubmersionoveronetidalcycleinthevariousscenariosof
sealevelriseasshowninTable4.Theanalysiswascarriedoutfor
thescenarioofthehighestfreshwaterdischarge(100m3/s).
Thetotalintertidalsubmersionareaincreasesfromavalueof
10km2forthepresentyeartoavalueof14km2inthe“2100
sce-nario.Therelationshipbetweenmeansealevelandfloodingarea
isnearlylinear.Furthermore,anincreaseinmeansealevel
8 L.Mills,J.Janeiro,A.A.S.Nevesetal./JournalofComputationalScience44(2020)101169
Fig.8.Histogramofflooddistributionareasasafunctionofthenumberofhoursofimmersionoveronetidalcycleinthevariousscenariosofsealevelrise.
Table4
Flooddistributionareasbasedonhoursofsubmersionoveronetidalcycleinthe variousscenariosofsealevelrise.
Class HoursofSubmersion Area(km2)
Present 2040 2070 2100 1 0–1.2 0.57 0.56 0.60 0.59 2 1.2–2.4 1.39 1.45 1.65 1.45 3 2.4–3.7 1.46 1.59 1.54 1.86 4 3.7–4.9 1.82 1.93 2.04 2.11 5 4.9–6.1 1.19 1.63 1.69 1.86 6 6.1–7.3 1.29 1.46 1.97 2.17 7 7.3–8.6 0.81 1.03 1.18 1.38 8 8.6–9.8 0.82 0.99 0.97 1.22 9 9.8-11 0.63 0.54 0.72 0.85 10 11–12.2 0.24 0.24 0.31 0.44 Total 10.23 11.41 12.67 13.93
upinfloodingclass,meaningahigheramountoflandwillbe
sub-mergedforalongerperiodwithrespecttosealevelrise.Theflood
distributionmaps(
Fig.9)demonstratethatareaslocatedfurtherdownstreamthe
estuaryarenotasaffectedbymeansealevelrisecomparedtoareas
locatedupstream.Theevolutionoftheinundatedareaismore
pro-nouncedinthenorthmarginoftheinletnotconfinedbyVilaReal
deSantoAntónio.ThemarshesneartheterminalsectionofRibeira
doBelichealsosufferhighinundationbecauseofsealevelrise.The
presentstudydoesnothavebathymetrydataforlocationsfurther
upstreamthispoint,sonoconclusionscanbedrawnforthe
flood-ingareasupstream.Sincetheriverhasanarrowerupstreamprofile
theareasoffloodingareexpectedtobelower.
4. Discussion
Themodelsimulationsdemonstratedthatthemixingand
trans-portofsalinityaretheresultofacomplexdynamicinwhichthe
lowerportionoftheGuadianaEstuaryisdependentonboth
fresh-waterflowfromtheGuadianaRiverandthemeansealevelofthe
adjacentcoastalwaters.TheGuadianaEstuaryalternatesbetween
purelyfreshwaterwhenthereisahighdischargefromthe
Gua-dianaRiverandwaterofvaryingsalinitiesforlowflowsfromthe
river.Inthelattersituation,thetidedominatesthesalinitywitha
semi-diurnalpattern.
For lower freshwater discharge rates, salinity increased in
responsetoanincreaseinmeansealevelwithamorepronounced
changeinsalinityinthemainchannel.Forhigherdischarges,the
increaseinwatervolumeduetosealevelriseinducedahigher
penetrationoffreshwaterfromtheoutletandthereforeresulted
inlowersalinityvalues atthemouthoftheestuary.The
veloc-itytimeseriesshowedareductioninadvectivetransportatthe
mouthoftheestuaryduetomeansealevelrise.Saltistransported
mainlybydiffusion,intheinnermostregionswherethesignalof
thetideisnotdetected.Thisresultisconsistentwithothervalues
ofvelocityobtainedforthis region.Timeseriesresultsfor
loca-tionsfurtherupstreamtheestuaryrevealedanincreaseinsalinity
ofabout4inresponsetosealevelrise.Theincreaseinsalinityin
responsetoanincreaseinmeansealevelismorepronouncedcloser
tothecoastlinethanitisupstreamthemainchannel.Theincreaseis
alsomuchmorepronouncedforthebathymetryallowingflooding,
withanearlylinearrelationshipbetweensealevelriseand
Fig.9. Flooddistributionmapsforadischargeof100m3/sfortheaverageM2tideforthevariousscenariosofsealevelrise.Colorsrepresenttheratiobetweentimebelow waterlevelandtotallengthoftidalcycle.
extensionofsalinityfurtherupstreamtheestuarywhenthe
simu-lationswererunoverthepresentbathymetry.Changesinsalinity
fortheflood-permittingbathymetryshowtheimpactthe
topog-raphyhasonthehydrodynamicsandsalttransportoftheestuary.
Theresultsfromthespecificlocationsusedinthetimeseries
anal-ysiscanbeusedbycoastalmanagerstodetermineiftheincrease
insalinityatCastroMarimwillimpacttheecosystem,society,and
theeconomy.
Theresultsoftheflooddistributionareasforalltypesoftide
demonstratedanincreaseinsubmergedlandareaduetomeansea
levelrisewithanestimatedincreaseof4km2fortheentire
inter-tidalareaofthepresentsituation.Thereismostsignificantflooding
atEsteirodaCarrasqueiraandthemarshlandsurroundingRibeira
deBeliche,asshowninFig.9.Resultsoftheareasoffloodingcan
beutilizedbycoastalmanagersandlandownerstodeterminewhat
actionshouldbetakentomanagetheinevitablefloodingduetosea
levelrise.
5. Conclusion
Thisstudyusedanumericalmodeltoanalyzeseveralimpacts
of sealevel risein theGuadiana Estuary.Sealevel rise
projec-tionsobtainedfromthe5threportoftheIPCCforRCP4.5andRCP
8.5 scenariosfor theyears 2040,2070and 2100wereimposed
in the model to simulatethree differentscenarios of sealevel
rise. The parameters of the simulations allowedan analysis of
thevelocitymodulus,salinitydistribution,and floodingareasof
theGuadianaEstuary.Severalvariables suchastide,freshwater
flow, meansealevel andchanges in topographyof theestuary
10 L.Mills,J.Janeiro,A.A.S.Nevesetal./JournalofComputationalScience44(2020)101169
levelrise.Thehighnumberofsimulationsallowedforananalysis
ofseveralpossibilitiesofthefuturestateoftheGuadiana
Estu-ary.
Themodeldemonstratedthatanincreaseinmeansealevel
gen-erallyresultsin adecreaseinthevelocitymodulusofthemain
channeloftheestuary.Thisismostlikelyduetotheincreasein
depthof themainchannelas wellas anincreasein water
vol-umeassociatedwiththeriseinmeansealevel.Increasesinsalinity
aremostsignificantforcasesofalowfreshwaterflowandfora
morelikelybathymetry thatallowsfloodingof thesurrounding
areas.
Theresultsportrayanoverallincreaseinbothsalinityand
flood-ingareaintheGuadianaEstuarywithrespecttoanincreaseinmean
sealevel.Asthesetwo consequencesaredetrimentalto
estuar-ineecosystems,societyandtheeconomy,furtherresearchmustbe
donetofurtheranalyzetheimpactsofsealevelriseinthe
Guadi-anaEstuary.Onelimitationofthisstudyisthatatwo-dimensional
modelwasusedasopposed toa three-dimensionalmodel.It is
representativeofmostcasesoftheGuadianaEstuarybecauseit
isgenerallywell-mixed,butaspreviouslymentionedtheestuary
canbecomepartiallystratifiedinhighfreshwaterflowconditions.
Athree-dimensionalmodelwouldprovideamorecomplete
anal-ysisofthe hydrodynamicsof theestuary. Additionalnumerical
modelsshouldbesimulatedtoportrayamorequantifiable
rela-tionshipbetweensalinityandsealevelriseandtofurtherexamine
howthesalinitychangeswithrespecttootherparameters,suchas
overtides.
DeclarationofCompetingInterest
Theauthorsdeclarethattheyhavenoknowncompeting
finan-cialinterestsorpersonalrelationshipsthatcouldhaveappearedto
influencetheworkreportedinthispaper.
CRediTauthorshipcontributionstatement
LaraMills: Writing - original draft,Formal Analysis,
Visual-ization.JoãoJaneiro:Methodology,Investigation,Resources,Data
curation.AntonioAugustoSeppNeves:Software,Formal
Analy-sis.Flávio Martins:Conceptualization,Methodology,Validation,
Investigation,Datacuration,Writing-review&editing,
Supervi-sion,Projectadministration,Fundingacquisition.
Acknowledgements
This work was developed under the project PIAAC-Algarve
“IntermunicipalPlanforAdaptationtoClimateChangesofAlgarve”.
References
[1]M.Abbott,A.Damsgaarde,G.Rodenhuis,System21Jupiter,Adesignsys-tem fortwo-dimensionalnearly-horizontalflows,J.Hyd.Res.1(1973)1–28. [3]A.R.Carrasco,O.Ferreira,D.Roelvink,Coastallagoonsandrisingsealevel:a
review,Earth.Rev.154(2016)356–368,http://dx.doi.org/10.1016/j.earscirev. 2015.11.007.
[4]V.P.Chua,M.Xu,Impactsofsea-levelriseonestuarinecirculation:an idealizedestuaryandSanFranciscoBay,J.Mar.Syst.139(2014)58–67,http:// dx.doi.org/10.1016/j.jmarsys.2014.05.012.
[5]J.A.Church,P.U.Clark,A.Cazenave,J.M.Gregory,S.Jevrejeva,A.Levermann, A.S.Unnikrishnan,2013:Sealevelchange.Climatechange2013:thephysical sciencebasis,in:ContributionofWorkingGroupItotheFifthAssessment ReportoftheIntergovernmentalPanelonClimateChange,2013,pp. 1137–1216,http://dx.doi.org/10.1017/CB09781107415315.026. [6]J.A.Church,N.J.White,A20thcenturyaccelerationinglobalsea-levelrise,
Geophys.Res.Lett.33(1)(2006)94–97,http://dx.doi.org/10.1029/ 2005GL024826.
[7]J.Delgado,T.Boski,J.M.Nieto,L.Pereira,D.Moura,A.Gomes,R. García-Tenorio,Sea-levelriseandanthropogenicactivitiesrecordedinthe latePleistocene/HolocenesedimentaryinfilloftheGuadianaEstuary(SW
Iberia),Quat.Sci.Rev.33(2012)121–141,http://dx.doi.org/10.1016/j. quascirev.2011.12.002.
[8]A.B.Fortunato,A.Oliveira,E.T.Alves,Circulationandsalinityintrusioninthe Guadianaestuary(Portugal/Spain),Thalassas18(2)(2002)43–65,December 2015.
[9]E.Garel,L.Pinto,A.Santos,Ó.Ferreira,Tidalandriverdischargeforcingupon waterandsedimentcirculationatarock-boundestuary(Guadianaestuary, Portugal),Estuar.Coast.ShelfSci.84(2)(2009)269–281,http://dx.doi.org/10. 1016/j.ecss.2009.07.002.
[10]T.W.Hilton,R.G.Najjar,L.Zhong,M.Li,Isthereasignalofsea-levelrisein ChesapeakeBaysalinity?J.Geophys.Res.Oceans113(9)(2008)1–12,http:// dx.doi.org/10.1029/2007JC004247.
[11]B.Hong,J.Shen,Responsesofestuarinesalinityandtransportprocessesto potentialfuturesea-levelriseintheChesapeakeBay,Estuar.Coast.ShelfSci. 104–105(2012)33–45,http://dx.doi.org/10.1016/j.ecss.2012.03.014. [12]J.Lopes,R.Neves,J.M.A.Dias,F.Martins,Calibrac¸ãoDeUmsistemaDe
modelac¸ãoparaOestuárioDo,4thSymposiumontheIberianAtlanticMargin (2003)4–6.
[13]MARETEC,MOHIDWater,RetrievedApril4,2019,from,2017http://wiki. mohid.com/index.php?title=MohidWater.
[14]F.Martins,P.Leitão,A.Silva,R.Neves,3Dmodellinginthesadoestuaryusing anewgenericverticaldiscretizationapproach,Oceanol.Acta24(S1)(2001) 51–62,http://dx.doi.org/10.1016/s0399-1784(01)00092-5.
[15]R.F.McLean,A.Tsyban,V.Burkett,J.O.Codignott,D.L.Forbes,N.Mimura,V. Ittekkot,CoastalzonesandMarineecosystems.Climatechange2001: impacts,in:AdaptationandVulnerability,pp.343–379,Cambridge,UK,2001. [16]R.J.Nicholls,N.Marinova,J.A.Lowe,S.Brown,P.Vellinga,D.DeGusmão,R.S.J. Tol,Sea-levelriseanditspossibleimpactsgivena“beyond4◦Cworld”inthe twenty-firstcentury.PhilosophicalTransactionsoftheRoyalSocietyA: mathematical,PhysicalandEngineeringSciences369(1934)(2011)161–181, http://dx.doi.org/10.1098/rsta.2010.0291.
[17]N.I.Pontee,B.A.Hamer,P.D.Turney,Sustainablefloodriskmanagementin Europeanestuaries,Proceedingsofthe1stInternationalConferenceon CoastalManagementandEngineeringinMiddleEast(ARABIANCOAST2005) (2005)(November).
[18]M.C.Quesada,J.García-Lafuente,E.Garel,J.DelgadoCabello,F.Martins,J. Moreno-Navas,Effectsoftidalandriverdischargeforcingsontidal propagationalongtheGuadianaEstuary,J.SeaRes.146(January)(2019) 1–13,http://dx.doi.org/10.1016/j.seares.2019.01.006.
[19]A.C.Ross,R.G.Najjar,M.Li,M.E.Mann,S.E.Ford,B.Katz,Sea-levelriseand otherinfluencesondecadal-scalesalinityvariabilityinacoastalplainestuary, Estuar.Coast.ShelfSci.157(2015)79–92,http://dx.doi.org/10.1016/j.ecss. 2015.01.022.
[20]D.M.R.Sampath,T.Boski,C.Loureiro,C.Sousa,Modellingofestuarine responsetosea-levelriseduringtheHolocene:applicationtotheGuadiana Estuary-SWIberia,Geomorphology232(2015)47–64,http://dx.doi.org/10. 1016/j.geomorph.2014.12.037.
[21]D.M.R.Sampath,T.Boski,P.L.Silva,F.A.Martins,Morphologicalevolutionof theGuadianaestuaryandintertidalzoneinresponsetoprojectedsea-level riseandsedimentsupplyscenarios,J.Quat.Sci.26(2)(2011)156–170,http:// dx.doi.org/10.1002/jqs.1434.
[22]W.J.Wiseman,E.M.Swenson,J.Power,TrendsinLouisianaEstuariesSalinity, Estuaries13(3)(1990)265–271.
LaraMillsjustcompletedamaster’sinsciencedegreein MarineandCoastalSystemsattheUniversityofAlgarvein Faro,Portugal.SheisoriginallyfromtheUnitedStatesbut hasbeenlivinginPortugalthepastthreeyearsto com-pletethedegree.Shereceivedherbachelor’sdegreein mathematicsattheUniversityofVermontin2013.
JoãoJaneirocompetedabachelor’sinsciencedegreein OceanographyattheUniversityofAlgarvein2004and laterreceivedhismaster’sinsciencedegreeinWaterand CostalManagementbytheGeophysicalInstituteinBergen UniversityinNorwayin2006.Heworkedasamodeler intheScottishEnvironmentalProtectionAgencyfortwo yearsandin2009hebeganaPhDinMarineSciences spe-cializinginNumericalModelling.