HAL Id: insu-01267158
https://hal-insu.archives-ouvertes.fr/insu-01267158
Submitted on 24 Apr 2017
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of
sci-entific research documents, whether they are
pub-lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Sediment transport induced by tidal bores. An
estimation from suspended matter measurements in the
Sée River (Mont-Saint-Michel Bay, northwestern France)
Lucille Furgerot, Dominique Mouazé, Bernadette Tessier, Laurent Perez,
Sylvain Haquin, Pierre Weill, Alain Crave
To cite this version:
Lucille Furgerot, Dominique Mouazé, Bernadette Tessier, Laurent Perez, Sylvain Haquin, et al..
Sedi-ment transport induced by tidal bores. An estimation from suspended matter measureSedi-ments in the Sée
River (Mont-Saint-Michel Bay, northwestern France). Comptes Rendus Géoscience, Elsevier Masson,
2016, Coastal sediment dynamics, 348 (6), pp.432-441. �10.1016/j.crte.2015.09.004�. �insu-01267158�
Stratigraphy,
Sedimentology
Sediment
transport
induced
by
tidal
bores.
An
estimation
from
suspended
matter
measurements
in
the
Se´e
River
(Mont-Saint-Michel
Bay,
northwestern
France)
Lucille
Furgerot
a,*
,b,
Dominique
Mouaze´
a,
Bernadette
Tessier
a,
Laurent
Perez
a,
Sylvain
Haquin
a,
Pierre
Weill
a,
Alain
Crave
ca
Universite´ deCaen,CNRSUMR6143MorphodynamiqueContinentaleetCoˆtie`re,24,ruedesTilleuls,14000Caen,France
b
Universite´ deLaRochelle,CNRSUMR7266LittoralEnvironnementsSocie´te´s,2,rueOlympe-de-Gouge,17000LaRochelle,France
c
CNRSUMR6118Ge´osciencesdeRennes,campusdeBeaulieu,35042Rennes,France
1. Introduction
Tidal currents in estuaries induce resuspension and transport of sediments. These processes are of great importanceforcoastalengineering,astheyhaveadirect impact on human facilities, such as dams, dykes and waterways. Over longer time scales, they are also responsiblefor thesediment infilling of estuaries ( Dal-rympleetal.,1992).Sedimentsuspensionandtransport are more intense during the peaks of flood and ebb
currents,thevelocityofwhichcanreachvaluesuptoafew meters per second. High suspended sediment concen-trations(SSC)aregenerally observedattheheadofthe estuary at the fresh/salt water interface. For instance,
Castaing(1981)measuredSSCof20g/Lduringafloodin theGaronneestuary.Thisturbiditymaximumzone(TMZ) hasbeen widelystudiedby in situ measurements(e.g.,
JiufaandChen,1998;UnclesandStephens,1993). Insometide-dominatedestuaries,tidalboresrepresent anotherpotentialmechanismforstrongsediment resus-pension.Thiswavepropagatingupstreaminchannelsis generatedbythewaterleveldifferencebetweenthefront oftherisingtideandtheriverlevelatlowflow,controlled bythefluvialdischarge.Thehydrodynamicsoftidalbores
C.R.Geoscience348(2016)432–441
ARTICLE INFO
Articlehistory:
Received4September2015
Acceptedafterrevision20September2015 Availableonline18January2016
HandledbySylvieBourquin
Keywords: Tidalbore Flowvelocity
Suspendedsedimentconcentration(SSC) Laboratorycalibration Directsampling ASMrod Sedimentflux Mont-Saint-MichelBay ABSTRACT
Tidalboresarebelievedtoinducesignificantsedimenttransportinmacrotidalestuaries. However,duetohighturbulenceandverylargesuspendedsedimentconcentration(SSC), themeasurementofsedimenttransportinducedbyatidalboreisactuallyatechnical challenge.Consequently,veryfewquantitativedatahavebeenpublishedsofar.Thispaper presentsSSCmeasurementsperformedintheSe´eRiverestuary(Mont-Saint-MichelBay, northwesternFrance) duringthetidalbore passagewithdirectandindirect (optical) methods.Bothmethodsarecalibratedinlaboratoryinordertoverifytheconsistencyof measurements, to calculate the uncertainties, and to correct the raw data. The SSC measurementscoupledwithADCPvelocitydataareusedtocalculatetheinstantaneous sedimenttransport(qs)associatedwiththetidalborepassage(upto40kg/m2/s).
ß2015Acade´miedessciences.PublishedbyElsevierMassonSAS.Thisisanopenaccess articleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/ 4.0/).
* Correspondingauthor.
E-mailaddress:lucille.furgerot@univ-lr.fr(L.Furgerot).
ContentslistsavailableatScienceDirect
Comptes
Rendus
Geoscience
ww w . sci e nc e di r e ct . com
http://dx.doi.org/10.1016/j.crte.2015.09.004
1631-0713/ß2015Acade´miedessciences.PublishedbyElsevierMassonSAS.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http:// creativecommons.org/licenses/by-nc-nd/4.0/).
havebeeninvestigatedinrecentstudies(Furgerot,2014; Huangetal.,2013;Simon,2013;Simpsonetal.,2004).The impactoftidalboresonsedimentresuspensionhasbeen estimatedbypunctualSSCmeasurements(Chansonetal., 2011;Fanetal.,2014;Furgerotetal.,inpress;Wolanski et al., 2004). Chanson et al. (2011) present the only quantitative estimation of sediment fluxes from ADV measurements.Unfortunately,thedeductionofSSCfrom ADV signalamplitudein onepointofthewatercolumn inducessignificantuncertainties.Inallcases,SSC measu-rements in highly turbidand turbulent flows remain a technical challenge. SSC can be measured by various techniques(acoustic,optical,sampling),whichallrequire a calibration in laboratory with sediment from the measurementsite(vanRijn,2007).
In this paper, we compare two methods of SSC measurement in loaded flow: the optical method and directsampling.WeapplythesemethodstoSSC measu-rementsduringatidalborepassageintheSe´eRiver (Mont-Saint-MichelBay,NWFrance).Afirstquantificationofthe bore hydrodynamics and SSC has been presented in
Furgerot et al. (in press). After a significant laboratory calibration step, we discuss the reliability of both measurement methods. Once corrected, the SSC data coupledwithvelocitymeasurementsallowustocalculate reliablesedimentfluxesduringtidalborepassages. 2. Measurementsite
2.1. TheMont-Saint-MichelBay:generalcontext
The Mont-Saint-Michel Bay (English Channel, NW France)isa500-km2depressionnestedbetweenBrittany andtheCotentinPeninsula(Fig.1A).Itischaracterizedbya C-type hypertidal regime (Archer, 2013), semi-diurnal with an insignificant diurnal inequality. The hypertidal rangethatreachesupto14mattheseawardentranceof thebayisduetothereflectionoftheincomingtidalwave from the Atlantic Ocean along the Cotentin Peninsula, which induces anamplified standingwave(Larsonneur, 1989).
Three mainriversflowintothesouth-eastern partof the bay, the Couesnon, the Se´lune and the Se´e rivers, forminganextensivetidalestuary.Thelargetidalranges inducehigh-energytidalcurrents(upto2m/s)controlling sedimentaryprocessesandchannelmigration(Lanierand Tessier,1998).Threedistinctestuarineareasaredefined basedontidalchannelmobility:
theexternalestuarywithhighlymobilechannels; themiddleestuarywithlowmigrationratechannels; theupperestuarywithchannelizedrivers(Fig.1A).
Tidal bores are observed potentially in these three zones (Furgerot, 2014), depending on local channel morphology.
2.2. Measurementsite‘‘LeBateau’’
Tidal boremeasurementswereperformedin theSe´e River, about 15km upstreamof theestuarymouth (i.e.
Tombelaine,Fig.1A)onasitenamed‘‘LeBateau’’where maximum tidal range reaches 1.5m. The mean annual river discharge is less than 10m3/s, and never exceeds
20m3/sduringthewetseason.Biennialflooddischargeis
intheorderof50m3/s(Bonnot-Courtoisetal.,2002). Thegeneralmorphologyofthechannelisa570-m-long straightsectionbetweenelbowmeanders,and bordered bymarshbanks(Fig.1A).Topo-bathymetricprofileswere performedacrossthechannelusingatachometer,covering alongitudinaldistanceof40m.Thespacingof measure-mentpointswasapproximately2m.Atotalof140points wereusedtoreconstructthechannelmorphology(Fig.1B). Thebankfullchannelwidthisaround30m, andisfairly constant along thestretch with a longitudinalslope of 0.16% in average. The channel cross-section (Fig. 1B) revealssteepbanks(308)andaflat20-m-widebottom. Water depth at low tide is frequently lower than 1m (Fig.2).Bankfullconditionsarereachedduringhightides andhighwaterdischarges,withwaterdepthsof2.5m.
Inadditiontoitslocalchannelmorphology,theriver water depth at low tide, exclusively controlled by the fluvial discharge, is a critical parameter for tidal bore development and shape (Huang et al., 2013). Breaking boresareobservedwhenthewaterdepthislow,causing
Fig.1.A.Spotimage(2007)oftheestuarinesystemintheeasternpartof Mont-Saint-Michel Bay. The site of tidal bore measurements (‘‘Le Bateau’’)islocatedupstreamontheSe´eRiver(redsquare).Thethree distinctestuarineareas(1,2,3)aredefinedonthebasisoftidalchannel mobility.CP:CotentinPeninsula.B.3DreconstructionoftheSe´eRiver channelat‘‘LeBateau’’.Thewaterlevelcorrespondstoalowtidelevelfor alowriverflow(V1).
anaerationofthewave.Undularboresdevelopwithlarger waterdepths,andarecharacterizedbyasuccessionawave formswithawavelengtharound2mandasmoothsurface. Iffluvialdischargeistoolarge,thetidalwave israpidly attenuated in its upstreampropagation, inhibiting tidal boredevelopment.Thisphenomenonwasobservedonthe Se´eRiverduringafloodeventinJanuary2014,withapeak dischargeabove35m3/s.
The shape of the tidal bore is also sensitive to the evolution of the water depth during its upstream propagation. Generally, the tidal bore is undular over thewholerectilinearsection.Then,itevolvesprogressively towardabreakingboreattheendofthesection.
At‘‘LeBateau’’measurement site,tidal rangecan be consideredaslargeincomparisonwiththeriverdischarge, whichisinaverageverylow,sothatinadditiontothelocal channelmorphology,conditionsarefavorableforsteady tidalboredevelopment.
2.3. Sedimentcharacteristics
Thebackscatteredsignalofopticaloracousticprobes used to measure suspended sediment concentration stronglydependsonthesedimentparticleshape, compo-sition, and size (Moate and Thorne, 2012; Thorne and Meral,2008).Itisthereforecriticaltoproperlycharacterize thesedimentinsuspensionatthemeasurementsite.
The solid discharge of the three rivers entering the bay is almost negligible (Bonnot-Courtois et al., 2002; Larsonneur,1989).Thisimpliesthatthesedimentinfilling theMont-Saint-MichelBayisalmostexclusivelyofmarine origin, and composed of siliciclastic material reworked fromtheEnglishChannelseafloorby tidalcurrentsand waves,andofbiogeniccarbonates(mainlyshellfragments andredalgae).
Themeansedimentsizedecreasesprogressivelyfrom offshoretothemostinternalpartsofthebay.Inthemiddle toupperestuary,sedimentsarecomposedofsiltysandsto sandysilts,namedlocally‘‘tangue’’(BourcartandCharlier, 1959; Larsonneur, 1989; Tessier, 1993). According to particlesizeanalysis,themodeofthe‘‘tangue’’grainsize distributionis116
m
mintheouterestuary(atGrouindu Sud–Fig.1A)anddecreasesupstreamto80m
matthe‘‘Le Bateau’’ site, where d10 is 25m
m and d90 is 138m
m,indicating well-sortedsediments.The‘‘tangue’’ contains anaverageof50%ofcarbonates,mainlycomposedofshell fragmentsandforaminiferatests,andalmostnoclayand organic matter.Thiscompositioninducesheterogeneous and highly irregular particle shapes, from spherical to elongatedorflatshapes.
Thispeculiarsedimentissuspectedtohaveasignificant influenceon themeasurement methodsoftheSSC, and thus implies a rigorous calibration of the instruments (direct and indirect) using sediment sampledfrom the studysite.
3. Methods
Variousparametersweremeasuredinsituduringthe passageof30tidalbores(10fieldcampaignsfromJanuary 2011 to May 2012): salinity, temperature, sediment rheology and grain size, flow velocities in the three directions andSSC. Onlytheresultsof velocityand SSC measurementsobtainedduringonecampaign(May2012) are presented herein. This is the only campaign where turbid water samples were collected at four elevations across the water column. The arrangement of the instrumentsusedinthispaperispresentedinFig.2. 3.1. Suspendedsedimentconcentration:measurementsand calibration
Anopticalmethodandadirectsamplingmethodwere usedforSSC measurements.In ordertogetreliableand comparabledata,bothwerecarefullycalibratedusingtwo differentsystems.
3.1.1. OpticalmeasurementswithanASM(ARGUSsurface meter)
The ASMrod records thesignalreflectivity (Rs)on a
verticalprofileat2Hz,thanksto144opticalsensors(OBS type)spaced 1cm apart.TheconversionfromRs toSSC
values (defined as CASM) has been done through a
calibration function defined in the laboratory prior to the field survey. Because nephelometric measurements dependon thesize,on therefractionindex,and onthe shapeofthesuspendedsediment,thecalibrationhastobe madewithsedimentsfromthesurveysite.Suspensionsat differentconcentrations(Creal)have beenpreparedwith
sedimentsampledinsitu,andintroducedinaclosedtube circuit.TheASMsensorsrecordedthereflectivityvaluesin thecircuitfor5minat1Hz.Thecalibrationcurve(Fig.3A) hasbeenbuiltbyassociatingthemeanofthereflectivity values (Rs) measured by thesensors with the real SSC
(Creal).TheASMrodiscalibratedforCrealrangingbetween
0and30g/L.Thisrepresentsagoodcompromisebetween
Fig.2. Insitu measurementsetup in thechannelcross-section (‘‘Le Bateau’’locality,Se´eRiver):ASM:ArgusSurfaceMeter;ADV:Acoustic DopplerVelocimeter;ADCP:AcousticDopplerCurrentProfiler;ms.l.: meansealevel.
L.Furgerotetal./C.R.Geoscience348(2016)432–441 434
the measurement range and the resolution of the SSC measurements. When concentrations exceed 30g/L, the sensors give a saturation value. The calibration curve providesanequationCASM=f(Rs)
D
CASMwithuncertain-ty
D
CASM. This uncertainty comprises the uncertaintyrelated to ASM rod (reflectivity measurement) and the uncertainty relatedtothesedimentconcentration intro-duced in thecalibration apparatus(Creal). The ASM rod
manufacturergivesarelativeuncertainty:
D
RsRs
¼10 %
TheuncertaintyofCrealisrelatedtotheuncertaintiesof
the sediment mass (mreal)and of the volume of water
(Vcircuit)introducedinthecalibrationsystem:
D
Creal Creal 2 ¼D
mreal mreal 2 þD
Vcircuit Vcircuit 2 whichgives:D
Creal Creal ¼1 to 2%These calculated uncertainties are reported as error barsinFig.3A.
3.1.2. Directsamplingbypumping
BecauseASMopticalmeasurementsarelimitedbythe sensor saturation value (at 30g/L), and to check the
consistency of this indirect method, water from the channel wassampledduring thepassage of tidalbores using manual pumps. Four sampling points were posi-tionedatdifferentlevelsinthewatercolumn,justabove thebedand at 20,40 and60cm abovethebed. Atthe passageofthetidalbore,amaximumsamplingfrequency ofabout0.5Hzwasachieved.Tworisingtideswithtidal boreweresurveyedbyusingthissamplingpumpsystem, andabout800samplespertidewerecollectedandstored inhermeticplasticpots.
Generally,SSC is determinedfromwatersamples by filtration.Thismethodisfasttoprocess,butfineparticles maybelostdependingonthequalityofthemeshfilters. We used another protocol to measure the suspended sedimentconcentration,whichpreservesthesedimentfor further investigations (grain size distribution measure-ments and sediment composition observation under a binocularmicroscope).Thisincludedthreesteps: aweighingofthepotcontainingthesedimentandthe
water(P1);
aweighingofthepotandthesediment(P2)afterdrying
at358C;
aweighingofthemassoftheemptypot(P3).
Fromtheseweights,theSSC(Csample)iscalculatedusing
thefollowingformula:
Csample¼
P2P3
P1P2
D
CsampleFig.3.CalibrationofSSCmeasuringinstruments.A.CalibrationcurveoftheASMrodbasedonreflectivitymeasurementsofseveralrealSSC(Crealfrom0to 30g/L).B.Recirculatingcircuitusedtocalibratedirectsamplingsbypumps.C.Calibrationcurvesfordirectsamplingbypumps.
With
D
Csample,theuncertaintyrelatedtotheweighingofP1,P2,andP3isdefinedas:
D
Csample Csample 2 ¼D
msed msed 2 þD
Vpot Vpot 2 withD
m2sed¼
D
P22þD
P32 and msed the mass of drysediment,
D
V2pot¼
D
P12þD
P22 and Vpot the volume ofsampledwater.
Theuncertaintyonthemass(
D
P1,D
P2,andD
P3)is0.2g,andisrelatedtothecharacteristicsoftheweighingscale.
DCsample
Csample variesfrom4%forconcentrationsof30g/Lto40%
forconcentrationsof1g/L.Theuncertaintyissignificant forlowconcentrationsbecauseitessentiallydependson weighing.
ThismethodofSSCmeasurementisalsocalibratedin the laboratory in a second calibration system. The experimental setup is made of a closed recirculating circuitinwhichaknownvolumeofwaterandsedimentis introduced.Arecirculationpumpensuresacertainlevelof turbulence in the circuit, and a homogeneous water/ sedimentmixture(Fig.3B).Therecirculatingturbidwater issampledwithamanualpumpidenticaltothatusedon thefield,andSSC(Csample)isdeterminedusingtheprotocol
described previously. Three concentrations (Creal at 5,
20and50g/L)werechosenforthecalibrationofthepump (Fig. 3C). For each concentration, the experiment was repeatedfivetimes(withcompletecleaningofthesystem). During each experiment,five samples were taken. This resultedin25samplesperconcentrationvalue.
MeasuredSSCvalues(Csample)areplottedagainstreal
SSC values (Creal) (Fig. 3C). On average, Csample slightly
overestimates Creal. A correction equation (Csample,cor=
f(Csample))maybeproposed,butasignificantdispersionof
valuesaroundtheaverageisobserved.Avariabilityinthe sampleconcentration valuesofabout10%isdetermined graphically, which has to be added to the uncertainty relatedtotheweighingprocess.Theuncertaintyfromthe calibrationsystemisnegligible:DCreal
Creal <0:7%.This
signifi-cantuncertaintycomesprobablyfromthepumpsystem: thesedimentcanbelostduringthefillingofthepots
(underestimationofCsample),or;
thesedimentcansettleintothesamplingsystem(pump and pipes), which could distort the next sample (overestimationofCsample).
Theseuncertainties have not beenestimated experi-mentally.
3.2. Currentvelocitymeasurements
Velocity measurements were performed with an AcousticDopplerCurrentProfiler(ADCP)andanAcoustic Doppler Velocimeter (ADV), both mounted on a metal structureanchoredonthechannelbank(Fig.2).TheADV (Vector Nortek) was placed horizontally and looking upstream60cmabove thechannelbed. Itrecordedthe threecomponentsofthevelocity(u,
v
,w)atafrequencyof64Hz. The ADCP (Workhorse Sentinel) was placed just below the water surface, lookingdownward (Fig. 2). It recordedverticalvelocityprofilesdown tothesediment bed at a frequency of 2Hz, with a signal frequency of 1200kHz. The ADCP was configured according to the desired acquisition frequency, the measurement condi-tions, and the size of the vertical measurements cells. Duringthefieldsurvey,thewaterdepthwaslow(1m) andthecellsizewassetat5cm.Themanufacturergivesa relativeuncertaintyonthevelocitymeasurement(
D
u/u)of 0.3%.4. Results
Theresultspresentedinthis paperfocuson thedata obtainedduringthesurveyofanundulartidalborethat developedduringthehighspringtideof8May2012(the tidalrangeintheouterestuarywas12.85m).TheSSCand velocityevolutionsduringandafterthepassageofthetidal boresof7May2012at‘‘LeBateau’’havebeendiscussedin
Furgerotetal.(inpress)(samefieldcampaign).Herein,we presentaquantificationofsedimentfluxesduringthetidal borepassageintheSe´eRiverfrom(i)SSCmeasurements (ASM probe and direct sampling), and (ii) velocity measurements (ADVand ADCPprobes). We discussthe reliabilityandthelimitsofeachmethodinhighlyturbid and turbulent water, andemphasize on theimportance anddifficultiesoftoolcalibrationinsuchenvironments. 4.1. Suspendedsedimentconcentration
We compare theresults of the ASM rod (CASM) and
direct samplingcorrected (Csample)afterlaboratory
cali-brations(Fig.4).
Data are presented at four sampling depths (0, 20, 40 and 60cm above channel bed). Overall, the SSC evolution measured with the ASM rod follows the SSC evolutionmeasuredwiththepumpsamplingmethod.As describedinFurgerotetal.(inpress),ahighincreaseinSSC during the tidal bore passage (the first seconds) is measured atthebottom(Fig.4D).Thesaturationofthe ASMrodsuggestedSSCgreaterthan30g/L,confirmedby direct sampling withSSC values up to35g/L. After the passageofthetidalborefront,SSCgraduallyincreasesover the whole water column. Although the global SSC evolution is similar with the two methods, the ASM measurements (CASM) show a constant positive offset
comparedtothepumpsamplingmeasurements(Csample).
This difference appears constant over time and is not explained by the errors due to the instrumentation previouslyidentified(envelopecurveinFig.4).
4.2. Flowvelocities
A detailed description of the velocity evolution measured using anADV vectoron thetwo components uandwisprovidedbyFurgerotetal.(inpress).Hereinwe supplementtheseADVdatawithADCPmeasurementsthat provideverticalprofilesoflongitudinalvelocityu(Fig.5). The average velocity (u) on the 60cm of the water columnis0.3m/s(Fig.5B,profilea).Theriverflowisvery
L.Furgerotetal./C.R.Geoscience348(2016)432–441 436
lowcomparedtotheflowafterthetidalborepassagewith u¼1 m=sintheupstreamdirection.TheADCPprofilebat the tidal borepassage (Fig. 5B) clearly shows that the longitudinal velocity inversion (from a downstream-directed to an upstream-directed flow) first occurs at thebottomofthewatercolumn.Thisinversionhasbeen recordedoverthefoursuccessivetidalboressurveyedon 7and8May2012.Theinversionthenoccursallacrossthe watercolumnalmostimmediatelyaftertheborepassage. The longitudinalvelocityis maximal just afterthetidal bore’spassage(upto2m/s)throughoutthewatercolumn. Then, theaveragedvelocitydecreases alongthevertical
profile, withsomefluctuations arounda mean valueof 0.7m/s(Fig.5B,profiled).
4.3. Sedimentfluxcalculation
Velocities(u)andSSC(CASM)dataareusedtoestimate
theinstantaneousadvectivesuspendedsedimentflux(qs).
Theformulausedforthiscalculationhasalreadybeenused by several authorsin comparable environmental condi-tions (Berni, 2010;Chansonet al.,2011; Desgue´eet al., 2011): qs¼ Z h 0 CASM u dz Xn i¼1
CASMðiÞ uðiÞ
whereqsanduarenegativeinthedownstreamdirection.
CASMisexpresseding/L(Fig.6A),thelongitudinalvelocity
component u is in m/s (Fig. 6B) and theinstantaneous sedimentfluxqsisinkg/m2/s(Fig.6C).Theuncertaintyon
thesedimentfluxcalculationcanbedeterminedaccording to the ADCP and ASM rod uncertainties calculated previously:
D
qs qs 2 ¼D
u u 2 þD
CASM CASM 2 Dqsqs isestimatedat10%.Themainsourceoferroristhe
ASMrodmeasurement.Despitethisuncertainty,weobtain anorderofmagnitudeofthesedimentflux.Alsonotethat the ASM rod saturates at 30g/L. There is therefore an underestimationoftheSSCandthesedimentfluxatthe tidalbore’spassage.
Beforethepassageofthetidalbore,thedownstream sediment flux is close to 0kg/m2/s. At the tidal bore’s
passage,aperiodofsignificantsedimentfluxoccursonthe bottom, with values reaching at least 404kg/m2/s
(Fig. 6C). During the 3 next minutes after the bore, the sedimentfluxinthemeasuredwatercolumnislower,with valueslessthan 101kg/m2/s.It thenincreases progres-sivelyinthewholemeasuredwatercolumnbetween4and 9minafterthepassageofthetidalbore,reachingmaximum values of 252.5kg/m2/s. Nine minutes after the bore, despiteSSCreachesitsmaximum,thesedimentfluxstartsto decrease as a consequence of the decrease in the flow velocity.
5. Discussion
WhileSSCevolutionassociatedwiththepassageofa tidalboreispreciselydescribedinFurgerotetal.(inpress), inthepresentpaper,weproposetodiscussthreemains points:
thereliabilityofSSCmeasurementsbydirectsampling; the comparison between ASM and direct sampling
measurements;
theinfluenceoftidalboresonsedimentfluxes. DirectsamplingsforSSCmeasurementsarecommonly used and generally allow the validation of indirect measuresofSSC(opticaloracoustic).Toourknowledge,
Fig.4.ComparisonbetweenASM(includingerrorenvelope)anddirect samplingmeasurementsatthesameelevations,i.e.at0(A),20(B),40(C) and60cmabovethebottom(D).Surveyof8May2012.
nostudyreportsacalibrationofthemethodtodetermine thevalidityofthedirectsamplingmeasures.However,our in situ data indicate that calibration is necessary for quantifyingaccuratelySSC.Inourcase,adifferenceof5g/L is measured between Csampleand CASM.In laboratory, a
slightoverestimationofCsampleisnotedinaverageandan
equation is proposed for corrected the in situ SSC measurements.Despitethiscalibrationandtheapplication oferrorbarsonthedata,somedifferencesbetweenCASM
andCsamplearerecordedinsitu.Weconsiderthatthemain
sourceoferrorcomesfromtheASMrod,becausethedirect samplings are checked and the potential sources of uncertainty have been identified and estimated. The ASM rod manufacturer gives two measurement ranges fortwodifferentparticlesizes:from0.005to5g/Lformud (d50=20
m
m) and from 0.05 to 50g/L for sand(d50=250
m
m).The‘‘tangue’’onthesiteisofintermediatesize(siltwithmodeat80
m
m).Becausetheprobehasbeen calibratedonalargerange(0to30g/L),abiasispossible duetotheparticlesize.Inaddition,accesstotherawASMdataislimited.Aftercalibrationoftherod,theinstrument providesonlydataing/Lduringfieldmeasurements.An error due toASM measurement is therefore difficultto identify.AstheoffsetbetweenCASMandCsampleisroughly
constant, in situ calibration of the ASM rod can be performedusingdirectsampling.
Despite theseerrors,it is interestingtocouplethese methods,whichshowsimilarSSC evolutionandpresent complementaryadvantages:
direct samplingis theonly waytodetermine reliable calibrationcurvesforacousticandopticalprobes,witha distinctionbetweenbackscatterandabsorptionphases ofthesignal.Directsamplingisalsoindependentofthe sediment heterogeneity, unlike indirectSSC measure-mentsthatcanbedependentongrainsize,shape and composition (Moate and Thorne, 2012; Thorne and Meral,2008);
theASMrodallowsaspace-timemeasurementwithan important resolution every centimeterat 2Hz, and a
Fig.5.A.Longitudinalvelocity(u)comparisonbetweenADVVectorandADCPWorkhorseSentinelat60cmfromtheinitialbedlevel.B.Verticalvelocity profiles(u)extractedfromADCPmeasurementsatfourdifferenttimes(a,b,candd,locatedonFig.5A).
L.Furgerotetal./C.R.Geoscience348(2016)432–441 438
monitoringofthesedimentbed’sevolution(erosionand deposition).Theprocessingofthismethodisrapidunlike direct sampling, which is time-consuming (drying, weighing).Inaddition,theASMrodisabletoperform measurementsinhighlyconcentratedflows,upto30g/L inthepresentcaseunlikemostavailableinstruments,i.e. from0to5g/LforanOBS-3(OpticalBackscatterSensor) andfrom0to20g/LforanABS(AcousticBackscattering System). However, large measurement ranges imply smalleraccuracies.
InparalleltoSSCmeasurements,ADCPmeasurements allowedustoexaminetheevolutionoftheflowvelocityin thewatercolumnatthepassageandsomeminutesaftera tidalbore.Ourdataclearlyevidencethattheflowinversion occursfirstonthebottombeforeitpropagatesinthewhole watercolumn.Ce´bile(2010)ontheGaronneRiver,aswell asBazin(1865)ontheSeineRiveralreadyreportedthat current reverses first on the bottom at the tidal bore passage. We suppose that this inversion causes a very significantshearstress onthesediment bedand subse-quenterosion.Onehypothesisisproposedinperspective ofthispublication.
From thesemeasurementsof SSC and velocity, sedi-mentfluxesarecalculated.Ourdatademonstratethat,on ourstudysite,upstream-directedsedimentfluxesbecome maximalinthewholewatercolumnabout8–9minafter thepassageofthetidalbore.Thishighlightsthepresence ofaturbiditymaximumzone(TMZ),whichisacommon featureintide-dominatedestuaries(e.g.,theSeineestuary,
Avoineetal.,1981,theGaronneestuary,Castaing,1981). Its development in the Mont-Saint-Michel estuary is quantified for the first time in our study. All the data
acquiredduringtheseveralsurveysonthetidalboresite ‘‘LeBateau’’evidencethatthesalinityfrontactuallyarrives morethan10minafterthetidalborepassageinaverage (Furgerot,2014).Thisfactdemonstratesthatflocculation processesduetothesaltwaterdo notexplaintheTMZ, unlikeinmostriverestuarieswheretheTMZdevelopsat the salt water/fresh water interface. We believe that a significantpartofthesedimentssupplyingtheTMZshould comefromthechannelbedandbankerosionduetothe tidalbore passage.After thepassage, theTMZstarts to develop,withvaluesofsedimentfluxupto25kg/m2/s,and thistransportpersistsoverafewminutes(about5minin ourcase).
Forthefirsttime,aquantificationofsediment fluxes duringahighspringfloodaccompaniedwithatidalboreis provided for a tidal channel of the Mont-Saint-Michel inner estuary. Very few quantitative data of sediment fluxeshave beenpublisheduntil nowin thishypertidal setting. Desgue´e et al. (2011)have calculated sediment fluxesduringspringtidesintheinnerestuary(Couesnon River)atthebeginningofthefloodveryclosetothebottom butonthebasisofdatacollectedonthetidalflat,notinto thechannel,andwithouttidalbore.Maximumvaluesof 1.1kg/m2/swerecalculated,whichis40 timesless than thesediment fluxes wehave calculated on thechannel bottomatthetidalborepassage.Ehrhold(1999)useda numerical model on the basis of data collected at the entrance of the estuary and calculated sediment fluxes greaterthan10kg/m2/sintheinnerestuaryforspringtide
conditions.Thesevaluesareaveragedoverthewholeinner estuaryandawholetidalcycle.Thisimpliesthatveryhigh sedimentfluxes,muchhigherthan10kg/m2/s,necessarily
occurintheinnerestuaryduringspringtideswhentidal
boresdevelop.Thisisconsistentwithourresults.Similarly,
Chansonetal.(2011)calculatedsedimentaryfluxesinthe GaronneestuaryduringatidalboreusingADV measure-ments at one point at the top of the water column. Downstream sediment flux was 1.5kg/m2/s before the
tidal bore. A few seconds after the passage, upstream sediment fluxes reached up to 100kg/m2/s and then
graduallydecreased to40kg/m2/sseveral minutes after
thebore.Takingintoaccountthehugedifferencesininitial conditionsbetweenoursiteandtheGaronneRiverestuary (fluvial flow and sediment fluxes, sediment grain size, channelsizeandmorphology),inadditiontothedifferent meansusedtomeasureSSC(ADVsignalinversionvsdirect samplingand ASMrod), weconsiderthat thevalues of sedimentfluxesassociatedwiththepassageofatidalbore obtained by Chanson et al. (2011) are in a range comparablewiththosewehavecalculatedinthisstudy. Although few values of sediment fluxes are finally available in estuaries where tidal bore can form, our comparative analysis tends to demonstrate that the developmentofatidalboreresultsinveryhighsediment fluxesinitiatedbytheborefrontandimpactonthebottom, oftheorderof10to40timeshigherthansedimentfluxes withouttidalbore.
Wewishtoquicklycomebacktothevelocityinversion, topresentourfirsthypothesisinperspectiveofthisarticle. Wesuggestthattheinversionoccursfirstatthebottom duetotwomainfeatures:
at tidal bore passage, hydrostatic pressure from the suddenriseofthefreesurfaceinadditiontoupstream velocities(topofwatercolumn)deflectsthedownstream riverflowtothemiddleofthewatercolumn;
atthesametime,closetothebottom,duetoathinner andsteeperboundarylayer,tidalflowvelocityishigher thanthefluvialone.
Thecombinedeffect ofthese twofeatures resultsin streamlinescontractionandaccelerationinthemiddleof thewatercolumnofriverflow.Thevelocityinversionis therefore starting at the bed, and is delayed at higher elevations.
6. Conclusion
Measurements of suspended sediment concentration andflowvelocitywereusedtocalculatesedimentfluxes duringthepassageofatidalborepassageintheSe´eRiver estuary (Mont-Saint-Michel Bay). From these measure-ments,threepointsaredevelopedinthispaper:
reliabilityofSSCmeasurementsbydirectsampling; comparison betweendirectsamplingand ASMrodfor
SSCmeasurements;
evolution ofsediment fluxwiththepassageof atidal bore.
Both ASM and direct sampling methods for SSC measurementswere calibratedinthelaboratory.Despite thelaboratorycalibration,adifferenceof5g/Latmaximum
was observed between both methods in the campaign results. Some hypotheses are proposed to explain this difference. This paper shows the difficultyof measuring reliableSSCevenwithdirectsampling,whichiscurrently considered as a reliable method. New tests should be performed to reduce the high uncertainty of the SSC measurements,includingamoreadvancedASMcalibration withdifferentrangesofparticlesize,andtheautomationof directsamplinginthefield.
Despite the uncertainties on theSSC measurements, sedimentfluxeswerecalculatedandcomparedwithother studiesconductedintheMont-Saint-MichelBaywithout tidalbore,andintheGaronneestuaryduringatidalbore.It shows thesignificantroleof thetidalboreon sediment transport byanimportantresuspension ofsedimentsat thepassageofatidalbore(Furgerotetal.,inpress).This importantresuspensionseemstocontributetothesupply of aturbiditymaximum zone(TMZ)highlightedforthe firsttimeintheMont-Saint-MichelBay.
Finally,weproposethattheinversionofvelocityonthe bottom is due to the difference in thickness of the boundarylayerbetweentheriverflowandthefloodflow belowthetidalborefront.Furtherprocessingofcurrent velocitydataisincoursetoconfirmthishypothesis.
Acknowledgements
This study is part of Lucille Furgerot PhD works (University of Caen), supported by the ANR project ‘‘Mascaret’’ (ANR-2010-BLAN-0911), whose coordinator, Prof.PierreLubin,iswarmlythanked,andbytheRegional CouncilofBasse-Normandie(CRBN).Wearegratefultoall the colleaguesof the ‘‘Mascaret’’ team and of the M2C research lab, as well as friends, who provided helpful assistanceforfieldsurveys.WewarmlythankJean-Yves Cocaign,theDirectorof‘‘LaMaisondelaBaie’’inthe Mont-Saint-Michel Bay, for thelogistical assistance he kindly supplied during the surveys. We also thank Teledyne InstrumentsfortheloanofseveralADCPincludingADCP WorkhorseSentinel.Finally,wethankPr.DaiduFanand Dr.BenoıˆtCamenenfortheirveryconstructivecomments. References
Archer,A.,2013.World’shighesttides:hypertidalcoastalsystemsin NorthAmerica,SouthAmericaandEurope.Sediment.Geol.284,1–25. Avoine,J.,Allen,G.P.,Nichols,M.,Salomon,J.C.,Larsonneur,C.,1981. SuspendedsedimenttransportintheSeineestuary,France.Effectof man-mademodificationsonestuary-shelfsedimentology.Mar.Geol. 40,119–137.
Bazin,M.,1865.Recherchesexpe´rimentalessurlapropagationdesondes. Me´moirespre´sente´spardiverssavantsa` l’Acade´miedessciences. Paris,France19,452–495.
Berni,C.,2010.Processusdemobilisationetdetransportdese´diments danslazonedede´ferlement(The`se).Univ.Grenoble,France. Bonnot-Courtois,C.,Caline,B.,L’Homer,A.,LeVot,M.,2002.Labaiedu
Mont-Saint-Micheletl’estuairedelaRance.Environnements se´di-mentaires,ame´nagementsete´volutionre´cente(Thebayof Mont-Saint-MichelandtheRanceEstuary:Recentdevelopmentand evo-lutionofdepositionalenvironments).Bull.CentreRech.Elf-Explor. Proc,Mem.26,256p.
Bourcart,J.,Charlier,R.,1959.Thetangue:a‘‘nonconforming’’sediment. Geol.Soc.Am.Bull.70,565–568.
Castaing,P.,1981.Letransferta` l’oce´andessuspensionsestuariennes (The`sed’E´tat).Universite´ deBordeaux,France.
L.Furgerotetal./C.R.Geoscience348(2016)432–441 440
Ce´bile,A.,2010.Lephe´nome`nedemascaret.Rapportdestagedemaster. Universite´ deBordeaux,France.
Chanson,H.,Reungoat,D.,Simon, B.,Lubin,P.,2011.High-frequency turbulenceandsuspendedsedimentconcentrationmeasurements intheGaronneRivertidalbore.Estuar.Coast.ShelfS.298–306. Dalrymple,R.W.,Zaitlin,B.A.,Boyd,R.,1992.Estuarinefaciesmodels:
conceptualbasisandstratigraphicimplications.J.Sediment.Petrol. 62,1130–1146.
Desgue´e,R.,Robin,N.,Gluard,L.,Monfort,O.,Anthony,E.,Levoy,F.,2011. Contributionofhydrodynamicconditionsduringshallowwater sta-gestothesedimentbalanceonatidalflat:Mont-Saint-MichelBay, Normandy,France.Estuar.Coast.ShelfS.94,343–354.
Ehrhold,A.,1999.Dynamiquedecomblementd’unbassinse´dimentaire soumisa` unre´gimeme´gatidal:exempledelabaiedu Mont-Saint-Michel(The`se).Universite´ deCaenBasse-Normandie,France. Fan,D.,Tu,J.,Shang,S.,Cai,G.,2014.Characteristicsoftidalboredeposits
andfaciesassociationsintheQiantangestuary,China.Mar.Geol.348, 1–14.
Furgerot,L.,2014.Proprie´te´shydrodynamiquesdumascaretetdeson influencesurladynamiquese´dimentaire.Uneapprochecouple´een canal etin situ (estuaire dela Se´e,baie duMont-Saint-Michel) (The`se).Universite´ deCaenBasse-Normandie,France.
Furgerot,L.,Mouaze´,D.,Tessier,B.,Perez,L.,2013.Suspendedsediment concentrationinrelationtothepassageofatidalbore(SeeRiver estuary,Mont-Saint-MichelBay,NWFrance).In:Proc.International conferenceofCoastalDynamics,Arcachon,France. pp.671–682. Furgerot,L.,Weill,P.,Mouaze´,D.,Tessier,B.,2016.Suspendedsediment
dynamicsinducedbythepassageofatidalboreinanupperestuary. In:IAS(InternationalAssociationofSedimentologists)SP. (special issue).
Huang,J.,Pan,C.H.,Kuang,C.P.,Zeng,J.,Chen,G.,2013.Experimental hydrodynamicstudyoftheQiantangRivertidalbore.J.Hydrodyn.25 (3),481–490.
Jiufa,L.,Chen, Z.,1998.Sediment resuspensionand implicationsfor turbidity maximum in the Changjiang Estuary. Mar. Geol. 148 (3–4),117–124.
Lanier,W.P.,Tessier,B.,1998.Climbingripplebeddinginfluvio-estuarine system;acommonfeatureassociatedwithtidaldynamics.Modern andancientanalogues.In:Alexander,C.,Davis,Jr.,R.A., Henry,V.J. (Eds.), Tidalites:ProcessesandProducts,61,SEPMSpec.Publ.,pp. 109–117.
Larsonneur,C.,1989.LabaieduMont-Saint-Michel.Bull.Inst.Ge´ol.Bas. Aquit46,1–75.
Moate,B.D.,Thorne,P.D.,2012.Interpretingacousticbackscatterfrom suspendedsedimentsofdifferentandmixedmineralogical composi-tion.Cont.ShelfRes.46,62–82.
Simon,B.,2013.Effectsoftidalboresonturbulentmixing:anumerical andphysicalstudyinpositivesurges(The`se).Universityof Queens-land,Australia.
Simpson,J.,Fisher,N.,Wiles,P.,2004.ReynoldsstressandTKEproduction inanestuarywithatidalbore.Estuar.Coast.ShelfS.60(4),619–627. Tessier,B.,1993.UpperintertidalrhythmitesintheMont-Saint-Michel Bay(NWFrance);perspectivesforpaleoreconstruction.Mar.Geol. 110,355–367.
Thorne,P.D.,Meral,R.,2008.Formulationsforthescatteringpropertiesof suspendedsandysedimentsforuseintheapplicationofacousticsto sedimenttransportprocesses.Cont.ShelfRes.28(2),309–317. Uncles,R.J.,Stephens,J.A.,1993.Thefreshwater-saltwaterinterfaceand
itsrelationshipto theturbiditymaximum intheTamarEstuary, UnitedKingdom.Estuaries16(1),126–141.
vanRijn,L.C.,2007.Principlesofsedimenttransportinrivers,estuaries, and coastalseas (update 2006).Aqua Publications, Blokzijl, The Netherlands.,www.aquapublications.nl.
Wolanski,E.,Williams,D.,Spagnol,S.,Chanson,H.,2004.Undulartidal boredynamicsintheDalyestuary,northernAustralia.Estuar.Coast. ShelfS.60,629–636.
