Citation for this paper:
Gibson, J.J., Yi, Y. & Birks, S.J., (2016). Isotope-based partitioning of streamflow in
the oil sands region, northern Alberta: Towards a monitoring strategy for assessing
flow sources and water quality controls. Journal of Hydrology: Regional Studies, 5,
131-148.
http://dx.doi.org/10.1016/j.ejrh.2015.12.062
UVicSPACE: Research & Learning Repository
_____________________________________________________________
Faculty of Social Science
Faculty Publications
_____________________________________________________________
Isotope-based partitioning of streamflow in the oil sands region, northern Alberta:
Towards a monitoring strategy for assessing flow sources and water quality controls
J.J. Gibson, Y. Yi, S.J. Birks
2016
© 2016 The Authors. Published by Elsevier B.V. This is an open access article
under the CC BY-NC-ND license (
http://creativecommons.org/licenses/by/4.0/
).
This article was originally published at:
http://dx.doi.org/10.1016/j.ejrh.2015.12.062
ContentslistsavailableatScienceDirect
Journal
of
Hydrology:
Regional
Studies
jou rn a l h om ep a ge :w w w . e l s e v i e r . c o m / l o c a t e / e j r h
Isotope-based
partitioning
of
streamflow
in
the
oil
sands
region,
northern
Alberta:
Towards
a
monitoring
strategy
for
assessing
flow
sources
and
water
quality
controls
J.J.
Gibson
a,b,∗,
Y.
Yi
a,b,
S.J.
Birks
a,caAlbertaInnovatesTechnologyFutures,3-4476MarkhamStreet,Victoria,BC,Canada bDepartmentofGeography,UniversityofVictoria,P.O.Box3060STNCSC,Victoria,BC,Canada cAlbertaInnovatesTechnologyFutures,360833rdStreetNW,Calgary,AB,Canada
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:
Received27October2015 Receivedinrevisedform 18December2015 Accepted23December2015 Availableonline14January2016 Keywords: Stableisotopes Hydrographseparation Groundwater Surfacewater Snowmelt Oilsands
a
b
s
t
r
a
c
t
Studyregion:ThisstudyisbasedontherapidlydevelopingAthabascaOilSandsregion, northeasternAlberta.
Studyfocus:Hydrographseparationusingstableisotopesofwaterisappliedtopartition streamflowsourcesintheAthabascaRiveranditstributaries.Distinctisotopiclabelling ofsnow,rain,groundwaterandsurfacewaterareappliedtoestimatethecontributionof thesesourcestostreamflowfromanalysisofmulti-yearrecordsofisotopesinstreamflow. Newhydrologicalinsightsfortheregion:Theresultsprovidenewinsightintorunoff genera-tionmechanismsoperatinginsixtributariesandatfourstationsalongtheAthabascaRiver. Groundwater,foundtobeanimportantflowsourceatallstations,isthedominant compo-nentofthehydrographinthreetributaries(SteepbankR.,MuskegR.,FirebagR.),accounting for39–50%ofannualstreamflow.Surfacewater,mainlydrainagefrompeatlands,isalso foundtobewidelyimportant,anddominantinthreetributaries(ClearwaterR.,Mackay R.,EllsR.),accountingfor45–81%ofannualstreamflow.Fairlylimitedcontributionsfrom directprecipitationillustratethatmostsnowandraineventsresultinindirectdisplacement ofpre-eventwaterbyfillandspillmechanisms.Systematicshiftsinregionalgroundwater tosurface-waterratiosareexpectedtobeanimportantcontrolonspatialandtemporal dis-tributionofwaterqualityparametersandusefulforevaluatingthesusceptibilityofrivers toclimateanddevelopmentimpacts.
©2016TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCC BY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Hydrographseparation based onstreamflowdatais one ofthemostwidelyused methodsfor quantifyingsurface water—groundwaterinteractionsatthereachtocatchmentscales(Kalbusetal.,2006).Typically,geochemicalorisotopicdata havebeenusedtotracechangesintheproportionofeventandpre-eventwatercontributionsduringstormsorsnowmelt events.AsnotedinarecentreviewbyKlausandMcDonnell(2013),thesestudieshaveforcedafundamentalre-examination oftheprocessesofwaterdeliverytostreams.Inparticular,theyhaverevealedahighproportionofpre-eventwaterinthe stormhydrograph,evenatpeakflow,(KlausandMcDonnell,2013).Dependingonthedistinctivenessofsoluteorisotope
∗ Correspondingauthor.
E-mailaddress:jjgibson@uvic.ca(J.J.Gibson).
http://dx.doi.org/10.1016/j.ejrh.2015.12.062
2214-5818/©2016TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/4.0/).
labellingandcatchmentpropertiesithasbeenpossibleinsomestudiestoresolvetwoorthreeseparatecomponentsofthe hydrograph,andtoinfermechanismsofrunoffgenerationsuchasgroundwaterridgingorvariablesourceareacontributions. Whileavarietyoftracerandnon-tracermethodshavebeenutilizedforhydrographseparation(seeGonzalesetal., 2009), stableisotopesofwaterhavetheadvantageofbeingincorporatedwithinthewatermoleculeandbeing mass-conservativeduringmixing.Stableisotopesofwaterareespecially usefulduetonaturallabellingofflow sourcesthat oftenarisesfromselectionandfractionationprocessesthatoccurinthewatercycle(Gat,1996).Severalexamplesinclude thedistinctlydepletedisotopicsignaturesnormallyassociatedwithsnow(andsnowpack)comparedtorainfallowingto temperature-dependentisotopicfractionationduringcondensationofatmosphericmoisture(Gat,1980),temporalisotopic variabilityinprecipitationwhich servestodistinguishindividualeventsfromlong-termaveragestypically reflectedin meteoricgroundwater(Gat,1996),selectiverecharge,whichmayaccentuatethedifferencebetweenprecipitationevents andmeteoricgroundwater,andevaporativeenrichmentofwetlandwatersorlakesthathaveresidedinsurfacestorage (Gibsonetal.,2015a).Evaporationfromsoilmayalsoleadtodistinctiveisotopiclabellingofshallowsoilwater,particularly inthearidzone(Gat,1981).
Hydrographseparationusingstableisotopesofwaterhasbeendemonstratedmainlyforhillslopesorsmallexperimental catchmentsovereventtimescales(Tetzlaffetal.,2015)withlessemphasisplacedonapplicationatlargerscalesorover seasonaltimeperiods(seeUhlenbrooketal.,2002).Thisstudyexploresuseofisotopehydrographseparationasanintegrated componentofstreamflowmonitoringinmeso-tomacro-scalepeatland-richcatchmentstobetterunderstandthephysical processescontrollingrunoffgeneration.Themainquestionsthatwewishtoanswerinclude:‘Howdopeatlandcatchments reacttosnowmeltandprecipitationevents?’,‘Whatisthetimingandproportionofvariouswatersourcestostreamflow onaseasonal,annualandinterannualbasis?’,‘Whatflowpaths,storageeffects,andrunoffmechanismsareimportant?’ and‘Arethereimportantregionaldifferencesinstreamflowresponse?’Oneoftheprimaryapplicationsweenvisionis betterunderstandingandpredictionofthetemporaldistributionofwaterqualityandcontaminantsinpeatland-dominated systems.Itisalsointerestingtoexplorewhetherinter-annualtime-seriesmonitoringmaybeusefulforassessingpotential changesinstreamflowdriversduetoclimateordevelopment-relatedimpacts.
Previousworkonstreamwaterchemistryincludinghydrographseparationstudieshavebeenconductedintheoilsands regionbySchwartz(1980)andSchwartzandMilne-Home(1982a,b).Basedonathree-yearrecordofmajoriontracersin fivemeso-scalecatchments,theypartitionedthehydrographintodirectprecipitation,groundwaterandmuskegwater,and weresuccessfulinshowingthatmuskeg(i.e.peatland)playsasignificantroleindeterminingthewatershedchemistry, inattenuatingrunoffduringspringmelt,andindilutionofwaterchemistryduringsummer,particularlybetweenrunoff events.Inaddition,theydeterminedthatgroundwaterplaysadominantroleduringwintercontrollingwaterquantityand quality.Notably,thesestudiesalsodemonstratedthatmeaningfulpartitioningcouldbeachievedintheregionoverseasonal tointer-annualtimeperiods.However,asthesestudiesusedgeochemicalratherthanisotopiclabellingforhydrograph separationtheydidnotattempttopartitionsnowmeltfromrainfall.
ThispaperoffersbasicconfirmationofsomeofSchwartz’sresultsfortheMuskeg,FirebagandSteepbankRivers,aswell assomeclarificationontheroleofdifferentrunoffcomponentsineachseasonandinter-annuallyfortheseandseveral additionalbasins(Mackay,Ells,ClearwaterandseveralstationsalongtheAthabascaRiver),basedondatasetsthatextend foroveradecadeinsomecases.Wepresentamethodologyforpartitioningofsnowmelt,rainfall,groundwaterandsurface waterusingstableisotopesofwaterandapplyittoinfersomeofthemajorprocessescontrollingrunoffformesoscale catchmentsaswellasfortheAthabascaRiverbothupstreamanddownstreamoftheoilsandsregion.Weidentifythe underlyingsourcesofrunoffandrunoffgenerationmechanismsthatproducespatialandtemporalvariationsinstreamflow inthewetland-muskegrunoffregime.Understandingrunoffgenerationintheregionisimportantasitisanessentialcontrol onwaterqualityandaquatichabitat,yetmaybechangingduetoongoingdevelopmentforoilsandsorduetoimpactof climaticchangesthatareknowntohaveaffectedpermafrostthawingandotherrunoffgenerationprocessesintheregion (Gibsonetal.,2015a,b).
1.1. Studyarea
ThestudyarealieswithintheAthabascaRiverbasin,Alberta,Canada.TheAthabascaRiverflowsnortheastover1,231km fromitsoriginsintheRockyMountainstothePeace-AthabascaDeltaandLakeAthabasca,draining156,000km2of land-scapevaryingfromsnow-cappedandglaciatedmountainstoagriculturalplains,borealforestandwetlands.Nodamsare constructedalongtheriverandconsumptivedivergencesofwateraresmallduetominimalagriculturaluse(Jasechkoetal., 2012).TheriverispartoftheMackenzieRiversystem,anditswaterseventuallyflowtotheArcticOcean.Thelowerreaches oftheAthabascaRivercoincidewiththeAthabascaOilSandsregion(AOSR).Here,approximately1%oftheriversannual flowisdivertedforuseasmake-upwaterforoilsandsminingandprocessingoperations(Canada,GovernmentofCanada, 2013).
Theclimateishighlyseasonalwithmonthlymeantemperaturesthatvaryfrom−19◦CinJanuaryto17◦CinJuly,with ameanannualtemperaturenear0◦C.Annualprecipitationis450mm,with60%fallingasrain.Reliefissubduedwiththe exceptionoflargeriverincisions.Fine-grainedsoils,incombinationwiththeclimate,haveresultedinformationofabundant wetlandsacrosstheregion.Ombrogenous(precipitation-fed)bogsandgeogenous(groundwater-fed)fens,whichtogether occupymorethan50%ofwatershedareasintheAOSR(Gibsonetal.,2015a),governhydrologyandinfiltrationatthesurface (Vittetal.,1994).Mineralsoiluplandsarealsocommoninthelower,incisedportionsofriverbasins.Peatlandsandmineral
Table1
Summaryofmonitoringstations,periodofrecordandsitecharacteristics.
StationID Stationname Daterangeofisotopesampling Area(km2) Discharge
Mean(m3/s) 1(m3/s)
07BE001 AthabascaR.atAthabasca 07/2002–05/2014 74,602 429 429
07DA001 AthabascaR.belowMcMurray 06/2002–05/2014 132,585 609 541
07DA0980a AthabascaR.atFirebagR. 08/2007–05/2014 n.a. n.a. n.a.
07DD011 AthabascaR.nearOldFort 05/2003–05/2014 156,000 768 750
07DA008 MuskegR. 06/2008–05/2014 1457 4.8 6.5 07DA006 SteepbankR. 04/2011–05/2014 1350 6.1 8.1 07DC001b FirebagR. 04/2011–05/2014 5990 29.4 25.4 07DB001 MackayR. 04/2012–05/2014 5570 16.9 28.3 07DA017 EllsR. 04/2011–05/2014 2450 7.1 12.1 07CD001 ClearwaterR. 04/2011–05/2014 30800 115.6 91.2
aWaterqualitystationonly.
bNotethatsampleswerecollectedslightlydownstreamatAB07DC0110.
soiluplandsareunderlainbyglacio-fluvialandglacio-lacustrinesedimentsthatcanexceed300mdepthinsomeareas overlyingburiedpaleochannels(AndriashekandAtkinson,2007).Permafrostintheregionissporadic,mainlyoccurringin bogsbutactivelydegrading(Vittetal.,1999).
RiverhydrologyintheAOSRisstronglyseasonal,withhighflowsassociatedwiththesnowmeltperiodinApril–May–June, andlowflowsassociatedwithice-coveredperiodsinNovembertoMarch.TheAthabascaRiverbelowFortMcMurraysustains flowsrangingfrom75to4700m3/s,withameandischargeof609m3/s(Canada,EnvironmentCanada,2015).Tributarypeak flowsintheAOSRaretypicallylessthan50m3/swithlowflowsoflessthan1m3/s(RAMP,2015).
RecentstudiesbyJasechkoetal.(2012),Gueetal.(2015),andGibsonetal.(2013)havedescribedthegeochemicaland isotopicsignaturesofnumerousspringsandbroaderseepagealongthelowerreachesoftheAthabascaRiver.Thesestudies revealtheroleofsalinebedrockformationwaterinsupplyingasmallproportion(upto3%)oftheflowtothelowerAthabasca River.Tributaries,asnotedbySchwartzandMilne-Home(1982b),arefedpredominantlybyshallowersourcesinglacial driftunits.OneexceptiontheynotedwasfortheSteepbankRiverwhichmayhavederived∼5%ofitsgroundwaterfrom deeperbedrockunits.
1.2. Samplecollectionandanalysis
TheAthabascaRiverandseveraltributariesweresampledonamonthlybasisaspartoftheLong-TermRiverNetwork monitoringprogramoperatedbyAlbertaEnvironmentalMonitoring,EvaluationandReportingAgencyanditspredecessors (Fig.1).Alistofsamplingstations,datesandbasiccharacteristicsofthewatershedsareshowninTable1.Watersamples werecollectedin30mLhigh-densitypolyethylenebottles(HDPE)followingstandardprotocolsforwaterqualitysampling (AlbertaEnvironment,2002).HDPEbottleshavebeenshowntobeveryeffectiveatpreventingisotopicfractionationfor periodsinexcessof1year(Spangenberg,2012).Snowdatawereobtainedfromaone-timeregionalsurveyofthesnowpack acrosstheoilsandsregionwhichincluded67isotopicanalysesofintegratedsnowpacksamples(Birksetal.,2014).Snowpack sampleswerefullymeltedinHDPEbagspriortobeingtransferredtoHDPEsamplebottles.Dataforrainfallwereobtained in2011–2012fromeventsamplingprogramssponsoredbytheCumulativeEnvironmentalManagementAssociationin twowatershedssituatedwithin70kmofFt.McMurray.SurfacewaterdatawereobtainedfromGibsonetal.(2015a)who reportedvaluesfor50lakesintheregionovera9-yearperiod.
Allisotoperesultswereanalyzedbyisotoperatiomassspectrometer,eitherattheUniversityofWaterloousinga Micro-massIsoPrimeDualInlet/GasChromatograph(pre-2009)oratAlbertaInnovatesTechnologyFutures,VictoriausingaThermo ScientificDeltaVAdvantageDualInlet/HDevicesystem.Inallcasesanalysesweremadewithin1yearofsamplecollection. Resultsarereportedin␦notationinpermil(‰)relativetoViennaStandardMeanOceanWater(V-SMOW)andnormalized totheSMOW-SLAPscalewhereSLAPisStandardLightArcticPrecipitation(seeCoplen,1996).Analyticaluncertaintyis estimatedtobebetterthan±0.1‰forı18Oand±1‰forı2H.
2. Theory
Massbalanceequationsarepresentedwhichdescribetherelativecontributionsofwatersourcestostreamflowunder asimplebatch-mixingmodelwithconservativetracers(Gibsonetal.,2000).Whileanapproachthataccountsforspatial andtemporalvariabilityinend-membersmaybemorerealistic(e.g.Harrisetal.,1995;OgunkoyaandJenkins1993),we currentlylackdetailedinformationtotracksystematicisotopicvariationsthatmaybeoccurringinthesourcewaters.Asa firstapproximationweapplythesimplebatchmixingmodeltoseparateinstantaneousstreamflowdischargeintoitssource componentsorend-members.
Fig.1.Mapshowinggaugingstationsintheoilsandsregionwhereisotopicsamplingwasconducted.Watershedboundariesfortributariesarealsoshown.
Forathree-componentsystem,theinstantaneousstreamflowdischargeQisequaltothesumofthecontributionsfrom thestreamflowsources(x1,x2,x3):
x1+x2+x3=Q (1)
Iftheisotopiccompositionofthewatersourcesisalsowell-constrainedthenadditionaltracerbalancescanbeconstructed. Inthecaseofı18Oandı2H,whicharemassconservative,themassbalanceequationsare:
x1ı181 +x2ı182 +x3ı183 =Qı18Q (2)
Fig.2.ı2H-ı18Oplotshowingtheisotopiccompositionofmajorstreamflowsourcesaswellastheiraveragevaluesmeasuredin2013.MWLismeteoric waterline.TwoareshownincludingtheGlobalMeteoricWaterLineofCraig(1961)(dashedline)andtheCanadianMeteoricWaterLineofGibsonetal., 2005(greyline).
whereı18
1 ,ı182 ,ı183 andı21,ı22,ı23aretheı18Oandı2Hofwatersourcesx1,x2,x3,respectively.
SolvingthesystemofEqs.(1)through(3)forthefractionalcontributionsofthecomponentsoftotalstreamflowyields: x1 Q =
ı18 Q −ı183 −ı2 Q−ı23ı18 2 −ı183 /ı2 2−ı23 ı18 1 −ı183 −ı2 1−ı23 ı18 2 −ı183 /ı2 2−ı23 (4a) x2 Q = ı18 Q −ı183 −ı2 Q−ı23 ı18 1 −ı183 /ı2 1−ı23 ı18 2 −ı183 −ı2 2−ı23 ı18 1 −ı183 /ı2 1−ı23 (4b) x3 Q = ı18 Q −ı181 −ı2 Q−ı21 ı18 2 −ı181 /ı2 2−ı21 ı18 3 −ı181 −ı2 3−ı21 ı18 2 −ı181 /ı2 2−ı21 (4c) Itshouldbenotedthatx1/Q,x2/Qandx3/QadduptounityasconstrainedbyEq.(1).AnanalyticalsolutiontoEqs.(4a–c) alsorequiresthatx1,x2,x3arenotcollinearinı2H-ı18Ospace.
3. Results
3.1. Isotopecharacteristics
Fourprimarystreamflowwatersourceswereidentifiedfortheoilsandsregion:snow,rain,surfacewaterand groundwa-ter.Theisotopiccompositionofsnow,rainandsurfacewaterwerecharacterizedbasedonwatersamplingprogramsinthe vicinityofFortMcMurray(Fig.2,Table2).Snowwasfoundtoplotclosetothemeteoricwaterline(MWL)forCanadagiven byı2H=8ı18O+8.5(Gibsonetal.,2005)andwassignificantlydepletedinheavyisotopesrelativetootherstreamflow sources.Similarpatternswerenotedpreviouslyforlong-termsnowsamplinginthelowerLiardValleybySt.Amouretal. (2005)andusedeffectivelyforsnowmelthydrographseparationinmesoscalebasins.Incontrast,summerrainwasfound tobeenrichedinheavyisotopes,plottingnearbutslightlybelowtheMWL.Surfacewatersaredistinguishedbysystematic evaporativeenrichment,plottingalongalocalevaporationline(LEL)givenbyı2H=5.20ı18O−50.6(Gibsonetal.,2015a). Thisisa9-yeardatasetbasedonsamplingin50lakesacrosstheregion.Waterinpeatlands(fensandbogs)subjectto evapo-rationhasasimilarevaporativesignatureplottingalongtheLEL,andasaresultarecountedassurfacewatercontributions.
Table2
Summaryofend-memberisotopiccompositionsandtheirvariability.
End-member N ı180 ı2H
Mean Max. Min. 1 Mean Max Min. 1
SN 67 −27.14 −25.60 −29.03 0.6 −206.7 −199.4 −220.5 3.5
RN 12 −4.07 −11.06 −16.98 1.9 −110.4 −87.3 −132.4 15.4
SW 50 −13.75 −10.35 −17.52 2.0 −121.1 −101.6 −143.2 10.7
GW 15 −20.75 −19.63 −22.56 0.8 −157.5 −148.5 −172.0 6.5
Note:SN—snow;RN—rain;SW—surfacewater;GW—groundwater.
Fig.3.Schematicshowingthetwo3-pointmixingscenariosusedinthehydrographseparation.Scenario1wasappliedduringthewinterandspring freshetperiodswhereasscenario2wasappliedduringsummer/fall.
StreamflowintributariesandtheAthabascaRivergenerallyhasanisotopiccompositionthatisintermediatebetweenthe
majorrunoffsources(Fig.2).
Theisotopiccompositionofregionalgroundwatersources,whichindirectlyreflectthemixingofsnowandrain,were estimatedastheinterceptofthewinterbaseflowrecessionandthemeteoricwaterlineinı2H-ı18Ospace.Regressionof winterstreamflowisotopicrecordswasmadebasedonseveralyearsofrecordinfivetributaries(MuskegR.[2008–2014], FirebagR.[2011–2014],SteepbankR.[2011–2014],EllsR.[2011–2014],andMackayR.[2011–2014])usingthemeteoric waterlineforCanada,givenbyı2H=8ı18O+8.5(Gibsonetal.,2005).Theregressionmethodwasusedinsteadofraw groundwaterdatatoensurethatthevaluesusedwereappropriatelyweightedtoreflectconditionsinthecatchmentsunder investigation.Maxima,minimaandaveragesarefoundtobesimilartotherangepreviouslyreportedforawidesurvey ofgroundwaterscollectedfromQuaternary,CretaceousandDevonianunitsintheoilsandsregion(Gibsonetal.,2013, 2015a;AndriashekandParks,2002),aswellasfromproprietarydatasetsownedbyAlbertaInnovatesTechnologyFutures.A verysimilaraveragevalueisobtainedastheinterceptoftheLELandtheMWL(ı18O=−21.11‰;ı2H=−160.4‰),which indicatesthemeansourceofinput(i.e.meanannualprecipitation+groundwater)tolakes.Groundwaterinourclassification includesallsubsurfaceflowcontributingtostreamflowandincorporatesshallowinterflowinpeatlandsaswellassurficial andbedrockaquifers.Itispresumedthatshallowinterflowinpeatlandswouldhavebeenclassifieddifferently,asmuskeg waters,bySchwartz(1980)andSchwartzandMilne-Home(1982a,b)basedongeochemicaltyping.
3.2. Modelsetupandoutputs
Asfourdistinctrunoff componentswereidentifiedand onlytwotracerswereused(Table2)somemethodological simplificationswererequired.Fortunately,duetothelimiteddurationofthesnowmeltperiod,mixinginthesystemcould beadequatelyapproximatedbasedontwotime-dependentthree-pointmixingscenarios.DuringAprilandMay,isotopic variationsduetomixingbetweensnowmelt,groundwaterandsurfacewaterweremodelledbasedonscenario1(Fig.3); fromJunetoOctoberisotopicchangesweresimulatedbasedonmixingbetweengroundwater,surfacewaterandsummer rainaccordingtoscenario2(Fig.3),andvariationsduringwintermonthsofNovembertoMarchweremodelledagainusing
Fig.4.Time-seriespartitioningsummaryforgroundwater-dominatedtributaries(a)MuskegR,2008–2014;(b)FirebagR.2011–1014;(c)SteepbankR., 2011–2014.
scenario1(Fig.3).Notethatthispartitioningapproachignorestheeffectofrainfallduringthesnowmeltperiod,butwas anecessarysimplificationtodealwithmixingoffourcomponentswithonlytwotracers.Ineffect,rainandsnowmixtures duringsnowmeltweretreatedasgroundwater—areasonableassumptiongiventhatthesewaterswillrarelyinteractoutside thegroundwaterenvironment.Junewasconsistentlytreatedasasummermonthinthis analysis,applyingscenario2, althoughitisalsopossibletotreattransitionmonthsaseithersnowmeltorsummermonthsdependingonwhenpeak snowmeltoccurs.
Bundlesofeightmixingcalculationswereusedforeachscenariobasedoncombinationsofpossiblemixingtriangles, choosingeithermaximumorminimumvaluesforeachend-member.Bundleswerethenaveragedratherthanusingsingle average-valuerunsinordertocapturepotentialuncertaintyintheend-members.Whileisotope-basedpartitioningwas conductedonmonthlybasis,andouranalysisandsummaries arebasedstrictlyupon themonthlyresults,forplotting purposesinFigs. 4–7wealsointerpolatedthepartitioningresultstomatchthedailyrecordofdischarge.Astep-wise interpolationmethodwasappliedusingfluxesofwaterratherthanpercentagesofthevariouscomponents.Similarresults
Fig.5.Time-seriespartitioningsummaryforsurface-water-dominatedtributaries:(a)ClearwaterR,2011–2013;(b)MackayR.2011–1014;(c)EllsR., 2011–2014.
werealsoobtainedintrialsusingalinearinterpolationmethodalthoughthestep-wiseapproachwasadoptedheredueto operationalsimplicity.Wealsopresentsummariesforthreedistinctperiodswithinthehydrologicalyear,consideredto runfromNovembertoOctober.Theseinclude:(i)theice-onperiod(November–March),(ii)thespringfreshet(April–May), characterizedbysnowmelt-drivenprocesses,and(iii)summer/fall(June–October),characterizedbymorevariableflows relatedtorainfall-drivenprocesses.
Uncertaintyforindividualsingle-valuemixingrunswasfirstestimatedbasedonthemethodofPhillipsandGregg(2001)
whichconsidersanalyticaluncertainty,samplesizeandstandarddeviationofend-members,aswellaspropagationoferrors duringmathematicalderivations.Basedonthisapproachuncertaintyforindividual(unbundled)runswasfoundtorange from±3%to±87%,withaveragevaluescloseto±27%.Thelargevariabilityinuncertaintycanbepartiallyattributedto mathematicalpropagationoferrors.Forbundledscenariosusedhereweestimateuncertaintybasedonstandarddeviation betweenrunstobesomewhatimproved:±9%forsnowmeltproportions,±12%forgroundwaterand±22forsurfacewater duringthefreshetandwinterperiods(i.e.Scenario1).Uncertaintyforsummerperiodswashigher:±26%forrain,±26%for groundwaterand±19%forsurfacewater(i.e.Scenario2).Greateruncertaintyinsummer/fallisprimarilyrelatedtoamore variableisotopicsignatureofrainthansnow.Overall,whiletheseuncertaintiesremainwithinausefulrangeforquantitative assessments,partitioningduringthesummerandfallneedstobeinterpretedmorecautiously.Thesystematicregional
Fig.6.Time-seriespartitioningsummaryforAthabascaR.atAthabasca,2002–2014.
Fig.7.Time-seriespartitioningsummaryforAthabascaR.belowFt.McMurray,2002–2014.
evolutioningroundwater/surface-waterratiospresentedlateronisstrongevidencethattheproportionsdeterminedare generallyconservativeandmeaningful.
3.3. Hydrographcharacteristics
Streamflowregime,thegeneralpatternofseasonalvariationinstreamflow,isinfluencedbywatersupply(e.g.snowmelt, rainfall,glaciermelt),waterlosses(e.g.evaporation)andstoragemodificationbylakes,wetlands,reservoirsandgroundwater (WooandThorne,2003).Smallstreamsintheoilsandsregionarecommonlyclassifiedaswetlandormuskegregime(see
Church,1974)wherelowreliefandaccumulationofpeatlandscreatesahighwaterretentioncapacityandresistancetoflow ofwaterwhichpromotesevaporation.Streamflowintheseareasistypicallyreducedduringthesummerperiodcompared tonon-wetlandregimes,andapronouncednival(snowmelt-generated)freshetmayoccurinlatespringwhenorganicsoils arestillfrozenandunabletoabsorbmeltwaterreleasedbysnow(Church1974).WooandThorne(2003)describedthe streamflowregimeintheAthabascaRiverbelowFortMcMurrayashavinganearlyhydrographriseduetosnowmeltin lowlands,followedbyasummerpeak,possiblysustainedbyglacierandhigh-elevationsnowmelt.Severeicejamsmayalso formalongtheAthabascaRiveratFortMcMurrayduringbreakupwhichcansignificantlyinfluencechannelstorageand waterlevels(AndresandDoyle,1984;Sheetal.,2009;Unterschultzetal.,2009).
Inter-annualvariationsinstreamflowclearlyrevealdifferentstylesofrunoffforthewatershedsexaminedhere(see totaldischarge,Figs.4–7),includingsharpandbroadpeaksinsomeyears.Severaldominantpeaksmayoccurinassociation witheithersnowmeltorsummer/fallstorms,withflashinesscontrolledalsobyfreezingconditions,antecedentmoisture andbyconnectivityofwetlandsandwaterbodies.Riverflowsaresomewhatmoresustainedherethanformorenortherly permafrostregionswhichmayceasetoflowduringwintermonths.
Table3
Seasonalandannualsourcewaterpartitioningsummaryforgroundwater-andsurface-water-dominatedtributariesintheoilsandsregion,northern Alberta.NotethattheMuskegR.isclassifiedasgroundwaterdominatedbasedonice-onflowconditions.Blankvaluesindicatethatthecomponentwas assumedtobezeroforthespecifiedtimeperiod.Smallnegativenumbersforgroundwaterandsnowmeltsuggestthattheyarelikelynotpresent.
River Timeperiod Flowcondition %GW %SW %SN %RN GW/SW
Groundwater-dominated
SteepbankR. Apr–May Freshet 51 27 17 1.89
Jun–Oct Summer/Fall 29 17 51 1.70
Nov–Mar Ice-on 69 26 −1 2.65
Annual Average 50 23 3 19 2.17
MuskegR. Apr–May Freshet 35 42 18 0.83
Jun–Oct Summer/Fall 23 39 34 0.59
Nov–Mar Ice-on 54 40 −1 1.35
Annual Average 39 40 3 14 1.00
FirebagR. Apr–May Freshet 49 29 17 1.69
Jun–Oct Summer/Fall 32 27 37 1.19
Nov-Mar Ice-on 51 35 9 1.46
Annual Average 44 31 7 14 1.42
Surface-waterdominated
MackayR. Apr–May Freshet 35 42 19 0.83
Jun–Oct Summer/Fall 14 46 36 0.30
Nov–Mar Ice-on 45 44 5 1.02
Annual Average 31 45 6 14 0.69
EllsR. Apr–May Freshet 8 77 10 0.10
Jun-Oct Summer/Fall -2 78 19 Undef.
Nov–Mar Ice-on 9 86 −1 0.10
Annual Average 5 81 2 7 0.06
ClearwaterR. Apr–May Freshet 38 52 4 0.73
Jun–Oct Summer/Fall 2 51 43 0.04
Nov–Mar Ice-on 29 69 −4.2 0.42
Annual Average 20 59 −1 14 0.34
Usingstableisotopesweendeavortotakeacloserlookattheunderlyingcausesofflowvariationsatspecificsitesinthe
followingsection.
3.4. Isotopictime-series
Isotopictimeseriesforı18OareshownforselectedstationsontributariesandtheAthabascaRiver(Fig.4–7).The
charac-teristicisotopicpatternnotedformostrecords/yearsisthatofarapiddeclineinisotopiccompositionduringsnowmeltwith agradualshifttowardsmoreenrichedvaluesinsummer/fallwhichextendsover-winter.Minorfluctuationsareattributed toraineventsinsummerandfall.Themostcommonvariationnotedisthatofaweakerordelayedspringmeltwhichresults inamoresubtleisotopicdepletion,oftenaccompaniedbyabroaderpeakinthehydrographitself.Examplesinclude2006 and2010forstationsalongtheAthabascaRiverand2010fortheMuskegRiver.Similartemporalvariationswerefoundfor ı2H(notshown)andwehavealreadyillustratedsystematicvariationsinı18Oandı2Hrelativetoflowsources(seeFig.2). 3.5. Partitioningofstreamflowcomponents
Partitioningresultsareshownforsixtributaries(MuskegR.,FirebagR.,SteepbankR.,ClearwaterR.,MackayR.,Ells R.;Figs.4and5)andtwostationsalongtheAthabascaRiver(Figs.6and7),wherethemajorsourcecomponents(rain, snow,surfacewaterandgroundwater)areshownascolour-codedbandsreflecting weightedcontributionstothetotal discharge.Overall,morethan609separatepartitioningdeterminationsweremadefor10stationsincludedinthemonitoring network,thevastmajority(95%)yieldingpositivepercentagesforendmembers.Thepartitioningresultsrevealasignificant groundwatercontributiontototalstreamflow(Tables3and4,Fig.8),evenduringhighflowepisodes.
Bothgroundwater-dominatedandsurface-waterdominatedtributariesareidentified.Forgroundwater-dominated sys-tems(SteepbankR.,MuskegR.,FirebagR.),groundwateraverages9–50%oftotaldischarge,rangingbetween23and32% duringsummer/fallwhensurfaceflowpathwaysaremostactive,51–69%duringwinter,and35–49%duringfreshet.Schwartz andMilne-Home(1982b)foundsimilargroundwater-dominatedconditionsintheMuskegandFirebagRiversbasedon hydrographseparationusingmajorions.Schwartzalsolookedatsmallersub-basinsoftheMuskegsuchasHartleyCreek andfoundevenhighergroundwatercontributions,sothesequantitiesareexpectedtobesomewhatvariableatsmaller scales.Schwartzdescribedthegroundwaterflowinthesetributariesasdominatedbyflowthroughglacialdrift(i.e. Quater-narydeposits),withasmallpercentagederivedfrombedrocksources.Overall,hepostulatedthatgroundwaterdischarge tosurfacewatersystemsintheuplandportionofthebasinswasdominatedbyshallowflowsystems,whereasitisderived increasinglyfromdeeperflowsystemswhichareintersectedinthelowerreachesofthebasinsasthestreamchannelbecomes moreincised.Surfacewateranddirectprecipitationaretheothermajorsourcesofrunoffingroundwaterdominatedsystems.
Athabasc
a R.
%GW
0 10 20 30 40 50 60 70 80 90 100% SW
0 10 20 30 40 50 60 70 80 90 100% RN
+ SN
0 10 20 30 40 50 60 70 80 90 100 Freshet Summer/Fall WinterGround
water-Dominated
Tribu
taries
%GW
0 10 20 30 40 50 60 70 80 90 100% SW
0 10 20 30 40 50 60 70 80 90 100% RN
+ SN
0 10 20 30 40 50 60 70 80 90 100Surface-Water
Domina
ted
Tribu
taries
%GW
0 10 20 30 40 50 60 70 80 90 100% SW
0 10 20 30 40 50 60 70 80 90 100% RN
+ SN
0 10 20 30 40 50 60 70 80 90 100Fig.8.Ternaryplotsshowingpartitioningresultsfor(a)AthabascaRiverstations,(b)groundwater-dominatedtributaries(SteepbankR.,MuskegR.,Firebag R.),and(c)surface-waterdominatedtributaries(ClearwaterR.,MackayR.,EllsR.).Contributionsarenormalizedto100%forplottingpurposes.Notein(a) thatsizeofdatapointsincreasesdownstream.
Table4
SeasonalandannualsourcewaterpartitioningsummaryforAthabascaRiverstationsintheoilsandsregion,northernAlberta.
Station Timeperiod Flowcondition %GW %SW %SN %RN GW/SW
Athabasca Apr–May Freshet 39 43 13 0.91
Jun–Oct Summer/fall 47 22 27 2.14
Nov–Mar Ice-on 45 44 6 1.02
Annual Average 45 34 5 11 1.32
FortMcMurray Apr–May Freshet 37 49 8 0.76
Jun–Oct Summer/fall 39 31 26 1.26
Nov–Mar Ice-on 41 49 4 0.84
Annual Average 39 42 3 11 0.93
Firebag Apr–May Freshet 39 45 10 0.87
Jun–Oct Summer/fall 31 34 31 0.91
Nov–Mar Ice-on 35 57 2 0.61
Annual Average 35 47 3 11 0.75
OldFort Apr–May Freshet 38 48 9 0.79
Jun–Oct Summer/fall 28 39 29 0.72
Nov–Mar Ice-on 36 55 2 0.65
Annual Average 33 48 2 12 0.68
Ouranalysisalsoextendstosurface-water-dominatedsystemswhichincludetheClearwater,MackayandEllsRivers.
Surfacewater,mainlyderivedfrompeatlandsandlakes,accountsfor45–81%oftotaldischargefromthesewatersheds,
rangingfrom46to78%duringsummer/fall,44–86%duringice-onperiod,and42–77%duringspringfreshet.Groundwater
isthesecondlargestsourceinthesewatershedsaccountingfor5–31%ofstreamflow.Surface-waterdominatedflowshave
alsobeennotedinsimilarlow-reliefterrainsuchaspeatlandwatershedsintheJamesBayLowlands(OrlovaandBrianfireun,
2014).
Groundwater/surface-waterratios(GW/SW;Tables3and4)areusefulfordistinguishingbetweenstations/seasonsthat aregroundwater-dominatedand thosethataresurface-waterdominated.Thegroundwater/surface-waterratiosinthe tributariesalsoappeartoinfluencetheratiosdeterminedalongthemainstemoftheAthabascaRiver.TheAthabascaRiver transitionsfromgroundwater-dominatedconditionsupstreamoftheoilsandsregionatAthabascatobalancedconditions nearFortMcMurrayandthenbecomessurface-waterdominatedinthedownstreamreachesneartheconfluencewiththe FirebagandatOldFort(Table4).Evolutionofthegroundwater/surface-waterratiosacrosstheregion(Fig.9)appearsto showthatflowfromsurface-waterdominatedtributariesbecomesprogressivelymoreimportantdownstream.
Oneofthemoststrikingfeaturesofourpartitioninganalysisisthatgroundwaterappearstobeasignificantcontributor tohighrunoff,includingbothsnowmeltandrainevents.Therehasbeenconsiderablediscussionanddebatewithinthe hydrologicalcommunityonmechanismsthatcanpromotegroundwater-dominatedrunoffduringpeakflows.Sklashand Farvolden(1979)suggestedthatgroundwaterincreasesduringstormeventswereduetogroundwaterridging,i.e., pre-cipitationcausesconversionofthenear-surfacetension-saturatedcapillaryfringeintophreaticwater,particularlyalong footslopesandvalleybottoms,leadingtoenhancedflowofgroundwatertostreams.Buttle(1994)alsosuggestedthat trans-latoryflowordisplacementofpre-eventwaterbyeventwatercouldbeimportant.Translatoryflowsoftenoriginatefrom alimitedportionofthedrainagebasin,asinthenearchannelareasorpermanentwetlandswheresurfacesaturationis maintained.Onemechanismthatiswidelyknowntooperateinwetland-dominatedareasissaturationoverlandfloworfill andspill(St.Amouretal.,2005;SpenceandWoo,2003),wherebyfrozenorunfrozensoilsmaybecomelocallysaturateddue tosnowmeltorrainevents,andoncedepressionstoragecapacityisexceeded,mayresultinflowoverlandorinmacropores. Inthiscasetherunoffistranslatoryinthatitwillcontainamixtureofeventwaterandgroundwater.Near-surfacepipes orrillshavealsobeenshowntobeimportantconduitsforpre-eventwatermovementfromhillslopestostreamsinsimilar wetlandterrain(Gibsonetal.,1993).Thesemechanisms,operatingsinglyorincombination,areareasonablehypothesisto explaintheincreasesingroundwaterdischargeduringbothrainandsnowmelteventsinthetributaries.
Surfacewatercontributionstototalstreamflow(asapercentage)werefoundtobegreatestinwinterforboth surface-andgroundwater-dominatedsystems.Whilewintergroundwaterdischargeisexpected,thedrainageofpeatlandwaters alsoapparentlyoccursthroughoutthewinteratallstations.Whilenotevidentinallsystems(exceptionsbeingtheEllsand MackayRivers),minimumcontributionsfromsurfacewatertendtooccurinsummer/fallduringtimesoflowantecedent moistureconditions.Surface-water-dominatedsystemssuchastheEllsandMackayRiversappeartobufferthiseffect, possiblyduetotheprevalenceoflakedrainageduringdryperiods.
Annualcontributionsfromdirectprecipitationsourcesrangedbetween0and7%forsnowmeltand7–19%forrainfall. Itislikelythatproportionsofthesesourcesmaybeunderestimatedattimesifeventsoccurredintheintervalsbetween sampling.Therelativecontributionsofgroundwater,surfacewateranddirectprecipitationareshowninFig.8.Overall,the dominantpatternisforwinterdischargetoliealongtheSW-GWaxis(i.e.directprecipitationisminimal),andfreshetand summer/fallperiodstendprogressivelytohavemoredirectprecipitationcontributions.Directprecipitationcontributions typicallyreflectfastrespondingrunoffgenerationmechanismssuchason-channelprecipitation,near-channelrunoffand overlandflow,althoughthelatterislikelytobeminorwithoutgroundwaterinteractioninpeat-dominatedterrain.Fast respondingsourcesalsotendedtodeclineslightlyatthelargerscale,accountingforlessthan50%ofstreamflowforthevast majorityofmonthlyrunsatstationsalongtheAthabascaR.ascomparedtoupto60%fortributaries.Thehighproportionof
Fig.9. Evolutionoftheisotope-basedgroundwater/surface-waterratiointheoilsandsregion.Notethattheinputofsurface-waterdominatedtributaries (ClearwaterR.,MackayR.,EllsR.)causesanapparentdeclineinratiosalongthemainstemoftheAthabascaRiver.
directprecipitationformanypeakflowshighlightstheimportanceofnear-channelareasforrunoffgenerationbothduring freshetandinsummer/fall.Contrastingresponsesarealsonotedforgroundwater-andsurface-waterdominatedsystems. Surface-waterdominatedsystemstendtohavelessgroundwaterinfluenceduringthesummer/fallbutincreasedinfluence duringwinter(Fig.8).
3.6. Monthlypatternsandinter-annualvariability
Thisisthefirststudytoourknowledgethathaslookedatmulti-yearsourcepartitioningsignalsacrossanetworkof mesoscaletributariesandalongalargeriversuchastheAthabascaR.Oneinterestingaspectofthisworkistoexamine thestabilityofthepartitioningresultsinmultipleyears.Forthegroundwater-dominatedtributaries,averystable inter-annualpictureemerges(Fig.10),with%groundwaterpeakinginlatefallandgraduallydecliningover-winterandintothe freshetandsummerperiods.Absolutequantitiesofgroundwaterareshowntomimicthetotalhydrograph,peakingduring thehighflowmonths(MaytoJuly),andreflectingtheroleofeventsintheoverallmechanismofgroundwaterdischarge. FortheSteepbankR.,theonlysignificantinter-annualvariabilitywasobservedinJanuaryandApril,althoughthespecific causeisunknown.Negativecorrelationsarenotedbetweenmeanairtemperatureand%groundwaterforthe groundwater-dominatedtributaries,suggestingthatcolderconditionsandicecoverfavorhighergroundwatercontributions.Notealso thelowerproportionsofgroundwaterintheEllsR.(<20%),asurface-waterdominatedtributary(Fig.10).Oneofthemost interestingobservationsismadefortheAthabascaRiverstationsincludingthestationbelowFortMcMurray(Fig.10),where wefindthatpercentageofgroundwaterpeaksatthesametimeasabsolutegroundwatercontribution,duringthehighflow period.Thisiscontrarytothegenerallyheldviewthatgroundwaterproportionstypicallypeakundericeinnorthern wetland-dominatedriversystems(seeGibsonandProwse,2002).Itisapparentfromourpartitioninganalysisthatgroundwaterplays animportantroleinrunoffgenerationthroughouttheyear.
Ells R.
Month
Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct
% G ro u nd water -10 0 10 20 30 40 50 m 3 /s 0 5 10 15 20
Steepbank
R.
MonthNov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct
% G roun d w at er 0 20 40 60 80 100 m 3/s 0 10 20 30 40 50
Athabasca
R.
McMurray
MonthNov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct
% G rou nd w a te r 10 20 30 40 50 60 70 80 m 3 /s 0 500 1000 1500 2000
Fig.10.Partitioningresultsfor%groundwaterandgroundwaterflow(m3/s)showinginter-annualvariabilitybymonth.EllsR.andSteepbankR.are exam-plesofsurface-andgroundwater-dominatedtributaries,respectively.NegativegroundwaterproportionsshownfortheEllsR.particularlyinAugust/Sept. suggestthatitislikelyabsentatthistime.
Surfacewater,whichislabelledinthisstudybytheuniqueisotopicsignaturethatwateracquireswhenexposedto evaporation,isalsoshowntobeanimportantflowregulatorinallseasons.Percentageofsurfacewaterisfairlystablefor tributariesoverthecourseoftheyearwiththeexceptionofsomesummermonths(particularlyJuly)forsurface-water dominatedtributariessuchastheEllsR.(Fig.11).Slightlyhigherproportionsareusuallyobservedinwinter,although thedifferencesaresubduedcomparedtogroundwater.Absolutequantitiesofsurfacewaterarelowestinwinterwhen hydrologicalpathwaysarefrozenordisconnected,highestinspringwhenthegroundismoresaturatedanddepression storageisgreater,withmorevariabilityduringthesummer/fallrelatedtocyclingofantecedentmoistureconditions.While %surfacewaterintheElls,MackayandClearwaterRiversmayaccountforcloseto100%ofdischargeinsomemonths, contributionsingroundwater-dominatedtributariesisoftenlimitedtolessthan40%,asrunofftypicallycontainsamixture ofsurfacewater,groundwaterandeventwater.Limitationsinthemixingmodelapproachusedarereflectedinsomemonths byproportionsofsurface-waterestimatedtobebelow0%orexceeding100%.Whilecomparativelymeaningfultheseresults needtobeinterpretedcautiously.
4. Discussion
Theapplicationofstableisotopetracerstodetectrunoffcomponents,combinedwithanunderstandingofthehydrologic settingofthelandscapeintheoilsandsregionenablesaconceptualmodelofrunoffgenerationmechanismstobeoutlined (Fig.12).Themajorprocessesidentifiedincludeon-channelprecipitationandnear-channeloverlandflow,whicharethought tobeimportantsourcesofdirectsnowmeltandrainfallrunoff.Inunsaturatedoff-channelareasinfiltrationcapacityisoften toohighfortheseprocessestooccur.Inwetlands,mixedwatersourcesaredeliveredtothestreamviafillandspill,which involveseventwaterraisingthegroundwatertableuntildepressionstorageissatisfiedandthenflowoccurs.Macroporeflow involvesflowofmixedwaterpredominantlythroughorganicsoilsandinterflow/returnflowsareshallowgroundwaters thatflowbetweenmineralsoilandpeat,contributingeithertofillandspillormacroporeflow.Thefillandspillprocess
Ells R.
Month
Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct
% S u rf ac e W a te r 0 20 40 60 80 100 m 3 /s 0 50 100 150 200
Steepbank R.
MonthNov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct
% S u rf ac e W a te r 0 20 40 60 80 100 m 3 /s 0 5 10 15 20 25
Athabasca R.
McMurray
MonthNov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct
% S u rf a c e W a te r 0 10 20 30 40 50 60 70 80 m 3 /s 0 500 1000 1500 2000
Fig.11. Partitioningresultsfor%surfacewaterandsurfacewaterflow(m3/s)showinginter-annualvariabilitybymonth.EllsR.andSteepbankR.are examplesofsurface-andgroundwater-dominatedtributaries,respectively.Whilecomparativelymeaningfulproportionsbelow0%andabove100%suggest limitationswithourquantitativemixingmodel.
Fig.12.Conceptualmodelofrunoffgenerationinwetland-dominatedtributariesintheoilsandsregion.Importantflowmechanismsareidentified. Notethaton-channelprecipitationandnear-channeloverlandflowproduceevent-dominatedrunoff,shallowrunoffcomponents(2–5)typicallyproduce mixturesofsurfacewaterandgroundwaterwhereasdeeprunoffcomponents(6–8)areexclusivelygroundwater-fed.
alsoappliestolargerbodiesofsurfacewaterincludinglakesandpondsthatareoftenconnectedbypermanentoutletsto thetributaries.Flowfromthesesourcesisisotopicallylabelledassurfacewaterduetoexposuretoevaporativeisotopic enrichment.Groundwatersourcesalsoincludeshallowflowindriftaquifersanddeepersources,includingbedrockaquifers orregionalgroundwater.
Ingeneral,theresidencetimeofwaterincreasesfromrunoffcomponent(1)throughto(8)(seeFig.12).Deeper ground-watersourcesalsotendtobemoreimportantlowerdowninthedrainagenetworkasthechannelsbecomemoreincised,in placeswithinbedrock.Ifgroundwaterridgingisoccurringinthewatersheds,thenitmustleadtointeractionwithorganic soils,otherwisewewouldnotseeincreasesinbothgroundwaterandsurfacewatercontributionsduringthefreshetor dur-ingrainevents.Translatoryflowsappeartodominaterunoff.Microtopography(hollowsandhummocks)inorganicterrain isimportantinthefillandspillprocessbecauseitpromotessubsurfaceinteractions(Freietal.,2010)andtranslatoryflows. Thefateofeventwater,ifitdoesnotfallinnear-channelareas,istobecomemixedwithgroundwaterifrecharged,orsurface waterifitlandsonsaturatedwetlandsandisexposedtoevaporation.
ThisstudyrefinesandextendstheinterpretationsofSchwartzandMilne-Home(1982a,b),particularlyfortheMuskeg andFirebagtributaries.SchwartzestimatedthegroundwatercontributiontotheMuskegR.during1976–1978torange between30and50%duringsummer,reducingto14–18%duringfreshet,and increasingto70–80%duringwinter.Our analysis,albeit foradifferentandextended timeperiod,suggestsanaveragegroundwatercontributionof34%during freshet,23%duringsummer,and54%duringwinter.SchwartzandMilne-Home(1982b)foundtheFirebagRivertocontain between10and20%groundwaterduringsummer,reducingtoaslowas2%duringfreshet,toashighas65%(butaveraging ∼30%)inwinter.Wefind32%groundwaterduringsummer,46%duringfreshetand51%duringwinter.Schwartzdeclinedto partitiontheSteepbankR.flowsasitderivedsignificantgroundwaterfrombedrock,anend-memberthathedidnotevaluate geochemically.FortheSteepbankRiver,ouranalysissuggeststhatgroundwater(includingbothdriftandbedrocksources) averages27%duringsummer,69%duringwinter,and49%duringthefreshet.
ThemaindifferencebetweenourassessmentandthatofSchwartzand Milne-Home(1982b)isthegreaterrolewe determinegroundwaterfluxestoplayduringthespringfreshet,whichisclearlyillustratedinFig.10.Thismayreflectbias intechnique,asourpartitioningapproachincludesshallowpeatlandgroundwater(whichisnotevaporativelymodified)in thegroundwaterclassification,whereasSchwartzandMilne-Homewouldhaveclassifiedthisasmuskegwater.Ourstudy alsousedthethree-componentsnowmelt-mixingscenario(Scenario1;Fig.3.)duringAprilandMayneglectedtheroleof rainfall,andmaythereforehaveledtoaslightoverestimationofgroundwatercontributionbyafewpercentduringthis time.Wefindadefinitivedecreaseintheproportionofgroundwaterduringfreshetascomparedtowinterandsummer/fall. However,moderatereductioninproportionsofgroundwatercombinedwithasignificantincrease instreamdischarge impliessignificantincreasesingroundwaterfluxesduringthefreshet.Meanwhileassumptionsmadeintheanalysisby Schwartz,includingthatprecipitationcontainednodissolvedions,seemstobeoversimplified.Anotherdifferencebetween ourassessmentandthatofSchwartzandMilne-Home(1982b)isthatwefindsimilarsourcepartitioningduringwinterin theMuskegandFirebagRivers,whereastheyfindaweakergroundwaterresponseintheFirebagRiver.Oneexplanation mightbethatflowpathsforgroundwateraredeeperintheMuskegRiverwhichwouldtendtoinfluencechemistrymore thanisotopiccompositionofstreamflow.Isotope-basedpartitioningalsoincludesshallowsubsurfacewaterasgroundwater, regardlessofthepathwaythatitentersthestream.Itisconceivablethatinsomecasesthiswatermayhavegeochemical propertiesthataremoresimilartopeatlandwaters.Ourapproachappearstocapturetheenhancedroleofgroundwater duringfreshetingroundwater-dominatedtributariesascomparedtosurface-waterdominatedtributaries,asillustratedin
Fig.10,whichdemonstratesencouragingsensitivityofthemethod.
Overall,ourmethodrevealsawiderangeinthegroundwatercontributionacrosstheregion.Wecanordertheriversin termsofgroundwaterdominanceasfollows:
fortributaries:Steepbank>Firebag>Muskeg>Mackay>Clearwater>Ells fortheAthabascaRiver:Athabasca>McMurray>Firebag>OldFort
GiventhatisotopicrecordsontheAthabascaRiveratFt.McMurraysometimesreachminimumvaluesaslateasJune orJuly,andthis occurswellafterlocalsnowmelthasconcluded,wepositthatsummerflowsalongthelowerreaches oftheAthabasca Rivermaybesupportedbysnowor glacialmeltin themountains,assuggestedbyWooandThorne (2003).ContributionofhighelevationrunofftostreamflowinJune/Julywouldresultinmorenegativeisotopevaluesfor thesemonthsandanoverestimationofgroundwatercontributioninthesummerperiod.Overall,glacialcontributionsare expectedtobeaminorsourceofflow,estimatedat0.8%ofdischargeduring2000–2007(Marshalletal.,2011).
Rasoulietal.(2013)showedthatflowintheAthabascaRiveratFt.McMurrayhasdeclinedbyroughly30%between1960 and2010,whichhashadalargeimpactonLakeAthabascawaterlevels.WooandThorne(2003)alsoreportedthatthe variabilityofdischargeintheAthabascaRiveratFt.McMurrayincreasedinthelatterhalfofthe20thcentury.Wepostulate thatthehighproportionofgroundwaterinthetotaldischargelikelyplaysasignificantroleinbufferingvariabilityand long-termtrendsinthetotaldischargeintheAthabascaRiver.Surface-waterdominatedtributariesthathavemorelimited groundwaterinputsmaybeespeciallysusceptibletoclimateordevelopmentimpacts.Assuch,wesuggestthatonefocus offutureresearchmightbetolookintolong-termtrendsintheindividualstreamflowcomponentstobetterunderstand thedriversofcurrentchange.ThiswouldlikelycomplementrecentstudiesofchangesinrunoffgenerationbyPetersetal. (2013).Anotheroneofthefundamentalimplicationsofthisresearchisthatthemethodmaybesuitableforexamining theunderlyingcausesofwaterqualitychanges,asthesemaybemoretightlycontrolledbytheoriginofthestreamflow andrunoffgenerationmechanismsthanthetotaldischarge.Forexample,thedistributionofsomeorganicspecies(such
asdissolvedorganiccarbon)maybecloselytiedtosurfacewatersourcesiftheyoriginatefrompeat,andsome(suchas naphthenicacids)maybemorecloselytiedtovariationsingroundwatercontributioniftheyoriginatefromcontactwith bitumenoranthropogenicsources.Monitoringofindividualrunoffcomponentsmayalsoaidincharacterizingdevelopment relatedimpactssuchasremovalofpeatland,forestsorgroundwaterabstractionthatmaydifferentiallyimpacttherunoff generationpathwaysacrosstheoilsandsregion.
5. Summary
• MonthlyisotopicrecordsofstreamflowarepresentedfortheAthabascaRiveranditstributariesintheAthabascaOilSands region,northeasternAlberta,withrecordsdatingbackto2002forsomestations.
• Anisotopicdatabaseofsourcewatersincludingsnow,rain,groundwaterandsurfacewaterwasusedtoestimatethe proportionofeachcomponentinstreamflowduringfreshet,summer/fallandwinterperiods.
• Groundwater- and surface-water-dominatedsystems are identified.Groundwater-dominated tributaries include the Steepbank,Muskeg,andFirebagRivers,wheregroundwateraccountedfor39to50%ofannualstreamflow,and surface-water-dominatedsystems,mainlysustainedbydrainagefromlakesandpeatlands,whichincludetheClearwater,Mackay, andEllsRivers.
• Evolutionofgroundwatertosurface-waterratiosacrosstheregionrevealsanoverallincreaseinsurfacewatersources downstream.
• Streamflowsourcesareexpectedtobeimportantunderlyingcontrolsonwaterquality,andmayinfluenceclimateand developmentimpactsonstreamflowandotherprocessesinarearivers.
Acknowledgements
FundingwasobtainedviaanNSERCDiscoveryGranttoJJGandviaanAlbertaInnovatesTechnologyFutures(AITF) ProgramInvestmentGrant.WethankEmilyTaylor,AITFforGISsupport.WeespeciallythankPrestonMcEachern,Purlucid Technologiesforassistanceandforesightininitiatingthesamplingprogramin2002andJessicaDiMariaandColinCooke, AEMERAforhelptomaintainandexpandtheprogram.
AppendixA. Supplementarydata
Supplementary data associated with this article can be found, in the online version, at
http://dx.doi.org/10.1016/j.ejrh.2015.12.062. References
AlbertaEnvironment,2002.WaterQualitySamplingMethods.WaterMonitoringGroup,ComplianceBranch,RegionalServices,AlbertaEnvironment. Andriashek,L.D.,Atkinson,N.,2007.Buriedchannelsandglacial-driftaquifersintheFortMcMurrayregion,northeasternAlberta.AlbertaEnergyUtilities
Board/AlbertaGeologicalSurveyReportESR.2007–01(March2007),160pp.
Andriashek,L.D.,Parks,K.,2002.BaselineinvestigationsintothegroundwaterresourcesoftheAthabascaoilsands(insitu)area,NortheastAlberta. AlbertaEnergyandUtilitiesboard,AlbertaGeologicalSurveyEUB/AGSSpecialReport98,480pp.
Andres,D.D.,Doyle,P.F.,1984.AnalysisofbreakupandicejamsontheAthabascaRiveratFortMcMurray,Alberta.Can.J.CivilEng.11,444–458.
Birks,J.,Y.,Yi,S.,Cho,E.Taylor,J.,Gibson,2014.CharacterizingtheorganiccompositionofsnowandsurfacewateracrosstheAthabascaRegion:Phase2. OSRINReportNo.TR-64.47pp.
Buttle,J.M.,1994.Isotopehydrographseparationsandrapiddeliveryofpre-eventwaterfromdrainagebasins.Prog.Phys.Geogr.18,16–41.
Canada,EnvironmentCanada,2015.WatersurveyofCanadaarchivedhydrometricdata,AthabascaRiverbelowFt.McMurray,
http://www.ec.gc.ca/rhc-wsc.(accessed17.09.15.).
Canada,GovernmentofCanada,2013.Oilsandswatermanagement—factsheet.
https://www.nrcan.gc.ca/sites/www.nrcan.gc.ca/files/energy/pdf/eneene/pubpub/pdf/11-%20OS-factsheetWater%20Management-eng-February-2013.pdf; (accessed29.09.15.).
Church,M.,1974.HydrologyandpermafrostwithreferencetonorthernNorthAmerica.In:Proceedings,WorkshopSeminaronPermafrostHydrology. CanadianNationalCommittee,InternationalHydrologicalDecade,Ottawa,pp.7–20.
Coplen,T.B.,1996.Newguidelinesforreportingstablehydrogen,carbon,andoxygenisotope-ratiodata.Geochim.Cosmochim.Acta60,3359–3360.
Craig,H.,1961.Isotopicvariationsinmeteoricwaters.Science133,1702–1703.
Frei,S.,Lischeid,G.,Fleckenstein,J.H.,2010.Effectsofmicro-topographyonsurface-subsurfaceexchangeandrunoffgenerationinavirtualriparian wetland—amodellingstudy.Adv.WaterResour.33,1388–1401.
Gat,J.R.,1980.Theisotopesofhydrogenandoxygeninprecipitation.In:Fritz,P.,Fontes,J.C.H.(Eds.),HandbookofEnvironmentalIsotopeGeochemistry, 1.Elsevier,Amsterdam/Oxford/NewYork,pp.21–47.
Gat,J.R.,1981.Groundwater.In:Gat,J.R.,Gonfiantini,R.(Eds.),StableIsotopeHydrology:DeuteriumandOxygen-18intheWaterCycle,10.International AtomicEnergyAgency,Vienna,pp.223–240.
Gat,J.R.,1996.Oxygenandhydrogenisotopesinthehydrologicalcycle.Annu.Rev.EarthPlanet.Sci.24,225–262.
Gibson,J.J.,Prowse,T.D.,2002.Stableisotopesinriverice:identificationofprimaryover-winterstreamflowsignalsandtheirhydrologicalsignificance. Hydrol.Processes16,873–890.
Gibson,J.J.,Edwards,T.W.D.,Prowse,T.D.,1993.RunoffgenerationinahighborealwetlandinnorthernCanada.NordicHydrol.24,213–224.
Gibson,J.J.,Fitzgerald,D.,Aravena,R.,Price,J.S.,Maloney,D.,2000.Streamflowgenerationinahypermaritimebog-forestupland.Hydrol.Processes14, 2711–2730.
Gibson,J.J.,Edwards,T.W.D.,Birks,S.J.,Amour,St.N.A.,Buhay,W.,McEachern,P.,Wolfe,B.B.,Peters,D.L.,2005.Progressinisotopetracerhydrologyin Canada.Hydrol.Processes19,303–327.
Gibson,J.J.,Fennell,J.,Birks,S.J.,Yi,Y.,Moncur,M.,Hansen,B.,Jasechko,S.,2013.EvidenceofdischargingsalineformationwatertotheAthabascaRiverin thenorthernAthabascaoilsandsregion.Can.J.EarthSci.50,1244–1257,dx.doi.org/10.1139/cjes-2013-0027.
Gibson,J.J.,Birks,S.J.,Yi,Y.,Vitt,D.,2015a.Runofftoboreallakeslinkedtolandcover,watershedmorphologyandpermafrostmelt:a9-yearisotopemass balanceassessment.Hydrol.Processeshttp://dx.doi.org/10.1002/hyp.10502.
Gibson,J.J.,Birks,S.J.,Yi,Y.,2015b.Highertritiumconcentrationsmeasuredinpermafrostthawlakes,northernAlberta.Hydrol.Processes(accepted 24.06.15.).
Gonzales,A.L.,Nonner,J.,Heijkers,J.,Uhlenbrook,S.,2009.Comparisonofdifferentbaseflowseparationmethodsinalowlandcatchment.Hydrol.Earth Syst.Sci.13,2055–2068.
Gue,A.E.,Mayer,B.,Grasby,S.E.,2015.OriginandgeochemistryofsalinespringwaterintheAthabascaoilsandsregion,Alberta,Canada.Appl.Geochem. 61,132–145.
Harris,D.M.,McDonnell,J.J.,Rodhe,A.,1995.Hydrographseparationusingcontinuousopensystemisotopemixing.WaterResour.Res.31,157–171.
Jasechko,S.,Gibson,J.J.,Birks,S.J.,Yi,Y.,2012.QuantifyingsalinegroundwaterseepageintheAthabascaoilsandsregion.Appl.Geochem.27,2068–2076.
Kalbus,E.,Reinstorf,F.,Schirmer,M.,2006.Measuringmethodsforgroundwater-surfacewaterinteractions:areview.Hydrol.EarthSyst.Sci.10,873–887.
Klaus,J.,McDonnell,J.J.,2013.Hydrographseparationusingstableisotopes:reviewandevaluation.J.Hydrol.505,47–64.
Marshall,S.J.,White,E.C.,Demuth,M.N.,Bolch,T.,Wheate,R.,Menounos,B.,Beedle,M.,Shear,J.M.,2011.Glacierwaterresourcesontheeasternslopesof theCanadianRockyMountains.Can.WaterResour.J.36,109–134.
Ogunkoya,O.O.,Jenkins,A.,1993.Analysisofstormhydrographandflowpathwaysusingathree-componenthydrographseparationmodel.J.Hydrol. 142,71–88.
Orlova,J.,Brianfireun,B.A.,2014.SurfacewaterandgroundwatercontributionstostreamflowintheJamesBayLowland,Canada.ArcticAntarcticAlpine Res.46,236–250.
Peters,D.L.,Atkinson,D.,Monk,W.A.,Tenenbaum,D.E.,Baird,D.J.,2013.Amulti-scalehydroclimaticanalysisofrunoffgenerationintheAthabascaRiver, westernCanada.Hydrol.Processes27,1915–1934.
Phillips,D.L.,Gregg,J.W.,2001.Uncertaintyinsourcepartitioningusingstableisotopes.Oecologia127,171–179.
RAMP,2015.RegionalAquaticsMonitoringProgram,http://www.ramp-alberta.org/data/ClimateHydrology/hydrology/Hydrograph.aspx,(accessed 17.09.15.).
Rasouli,K.,Hernandez-Henriquez,M.A.,Dery,S.J.,2013.StreamflowinputtoLakeAthabasca,Canada.Hydrol.EarthSyst.Sci.17,1681–1691.
Schwartz,F.W.,1980.HydrogeologicalinvestigationofMuskegRiverbasin,Alberta.AlbertaOilSandsEnvironmentalResearchProgram,Report48,97pp. Schwartz,F.W.,Milne-Home,W.A.,1982a.Watershedsinmuskegterrain:1.Thechemistryofwatersystems.J.Hydrol.57,267–290.
Schwartz,F.W.,Milne-Home,W.A.,1982b.Watershedsinmuskegterrain:1.Evaluationsbasedonwaterchemistry.J.Hydrol.57,291–305.
She,Y.,Andrishak,R.,Hicks,F.,Morse,B.,Stander,E.,Krath,C.,Keller,D.,Abarca,N.,Nolin,S.,Tanekou,F.,Mahabir,C.,2007.AthabascaRiverIceJam formationandreleaseevents,2006and2007.J.ColdRegionsSci.Technol.55,249–261,SpecialIssueonRiverIce.
Sklash,M.G.,Farvolden,R.N.,1979.Theroleofgroundwaterinstormrunoff.J.Hydrol.43,45–65.
Spangenberg,J.E.,2012.Cautiononthestorageofwatersandaqueoussolutionsinplasticcontainersforhydrogenandoxygenstableisotopeanalysis. RapidCommun.MassSpectrom.26,2627–2636.
Spence,C.,Woo,M.K.,2003.HydrologyofsubarcticCanadianShield:soilfilledvalleys.J.Hydrol.279,151–166.
Amour,St.N.A.,Gibson,J.J.,Edwards,T.W.D.,Prowse,T.D.,Pietroniro,A.,2005.Isotopictime-seriespartitioningofstreamflowcomponentsin wetland-dominatedcatchments,lowerLiardRiverBasin,NorthwestTerritories,Canada.Hydrol.Processes19,3357–3381.
Tetzlaff,D.,Buttle,J.,Carey,S.K.,McGuire,K.,Laudon,H.,Soulsby,C.,2015.Tracer-basedassessmentofflowpaths,storageandrunoffgenerationin northerncatchments:areview.Hydrol.Processes29,3475–3490.
Unterschultz,K.,vanderSanden,J.,Hicks,F.E.,2009.PotentialofRADARSAT-1fortheMonitoringofRiverIce:resultsofacasestudyontheAthabasca RiveratFortMcMurray,Canada.J.ColdRegionsSci.Technol.55,238–248,SpecialIssueonRiverIce.
Uhlenbrook,S.,Frey,M.,Leibundgut,C.,Maloszewski,P.,2002.Hydrographseparationinamesoscalemountainousbasinateventandseasonal timescales.WaterResour.Res.38(6),1096,http://dx.doi.org/10.1029/2001WR000938.
Vitt,D.H.,Halsey,L.A.,Zoltai,S.C.,1994.TheboglandformsofcontinentalwesternCanadainrelationtoclimateandpermafrostpatterns.ArcticAlpine Res.26,1–13.
Vitt,D.H.,Halsey,L.A.,Zoltai,S.C.,1999.ThechanginglandscapeofCanada’swesternborealforest:thecurrentdynamicsofpermafrost.Can.J.For.Res.30, 283–287.