peer review report 1 on assessing the role of uncertain precipitation estimates on the robustness of hydrological model parameters under highly variable climate conditions

Zambezithecorrelationbetweenmonthlyrainfallandstreamflowisaboutr=0.25,butsimulation resultscanbequite accuratewithKGE’=0.90andaboveseeKlingetal.,2014,JournalofHydrology:RegionalStudies:“I

Journal of Hydrology: Regional

Studies

j ou rn a l h o m e pa g e :w w w e l s e v i e r c o m / l o c a t e / e j r h

Peer Review Report

Peer review report 1 on “Assessing the role of uncertain precipitation estimates on the robustness of hydrological model parameters under highly variable climate conditions”

1 Original Submission

2 Comments to Author:

regards,

DOI of the original article: http://dx.doi.org/10.1016/j.ejrh.2016.09.003

Zambezithecorrelationbetweenmonthlyrainfallandstreamflowisaboutr=0.25,butsimulation resultscanbequite accuratewithKGE’=0.90andabove(seeKlingetal.,2014,JournalofHydrology:RegionalStudies:“Impactmodellingof waterresourcesdevelopmentscenariosonZambeziRiverdischarge”)

ThereasonisthatinmanyAfricanbasinstherainfallatthestartoftherainyseasondoesnotgeneraterunoffasthesoil firsthastoreachaspecificwetness.Consequentlythereislowcorrelationbetweenrainfallandstreamflowtime-series.I suggesttheauthorssimplynevermentionthesixexcludedbasins(youdonotneedtodwellonthelowcorrelation)

Aprioriparametervalues:

Onpage5youdescribethedatasourcesforaprioriparameterestimationfortheLISFLOODmodel.However,lateron

inthepaperyouneverdiscussthesimulationresultswiththeaprioriparameters.Itwouldbeinterestingtoreportabout themodelperformanceinthetwoevaluationperiods(isitalwayspoorerthanwithcalibratedparametervalueswhen switchingcalibration/evaluationperiods?),whatperformancesub-componentsareaffected(r,beta,gamma)andwhatis therelationshipbetweenaprioriparametervaluesandcalibratedvalues?Thesimulationwithaprioriparameterswould serveasagoodbaselineagainstwhichallothersimulationresultscouldbecompared

Kling’spersonalcommunication(2012):

Onpage15(line348–350)youciteKling’spersonalcommunicationforclassificationofKGEvaluesintogood/medium/etc Hereisanexcerptfromtheoriginale-mailtoVeraThiemig(18.10.2012)servingasthereference:

[—e-mailexcerptstart—]

Giventheabovethreepoints,Isuggestyouusemuchstrictervaluesforclassification,suchas:

1.00>KGE>0.75:good

0.75>KGE>0.50:intermediate

0.50>KGE:weak

Iexpectthatforthe“weak”simulations,themostlikelycausearebiasedprecipitationinputs,andnotsomuchthe calibrationofthehydrologicalmodel.Thiswouldshowupinpoorbetavalues,butgoodvariabilityratios(ifyoubasethe variabilityratiooncoefficientofvaration).Therefore,IstronglysuggestthatyoubasetheevaluationnotonlyonKGE,but alsoonthethreesub-components,asthisisagreatwayoflearningthecausesforpoormodelperformance:poortiming andshape−>poorcorrelation,poormassbalance−>poorbiasratio,poordistributionofflows−>poorvariabilityratio [—e-mailexcerptend—]

AvalueofKGE=0.75isnotyieldinga hydrographsimulationwhichlooksperfectlyaccurate.Therefore itmightbe worthwhiletoincludeanupperclassKGE>0.9named“excellent”.IwouldalsoliketopointoutthatKGEisjustasummary measureanditdependsonthemodelapplicationwhatsub-componentsareimportant.Forexample,forassessinglong-term reservoirinflowsthebiasisimportant,butcorrelationisnotsoimportant.Forfloodforecastingapplicationsthebiasisnot importantbutcorrelationisveryimportant.Forassessinglong-termenergygenerationatrun-of-riverhydropowerplants thebiasandvariabilitytermsareveryimportant(flowdurationcurve),butnotcorrelation.Insummary,youhavetoknow theapplicationofthemodel,beforeareasonableclassificationintogood/medium/etc.performanceisfeasible.Whatisthe intendedapplicationoftheLISFLOODmodel?

Fig.4andFig.9:

Iunderstandthedotsshowninthefigures,butafterseveralminutesofthinkingIhavetoadmitIamstillnotsurehowto preciselyinterpretthebars.AndIamsuresomeofthereaderswillhavesimilardifficulties.Pleasegiveabetterdescription

inthecaptionandtext

Specificcomments:

“anon-stationaryhydrologicalmodel”

Themodelisnotnon-stationary,probablyyoumean“non-stationary”climate?However,seealsomygeneralcomments aboveabouttheuseoftheterm“non-stationaryclimate”

Abstract,firstparagraph:

“LISFLOODhydrological,and”−>“LISFLOODhydrologicalmodel,and”

page1,line12:

Isuggesttoremove“However”atthestartofthesentence

page1,line14,15:

Isuggesttoremove“Ontheotherhand”atthestartofthesentence

page3,line68:

“reliabletime-seriesofprecipitation”

Iwouldremovetheword“reliable”whenreferringtosatellite-basedprecipitationdata

page3,line71:

Inthereferencesyoushouldalsocite:

Cohen-Liechtietal.,2012.HESS:Comparisonandevaluationofsatellitederivedprecipitationproductsforhydrological modelingoftheZambeziRiverBasin

page4,line92:

“variabilityinwill”−>“variabilitywill”

page7,line162:

ForTRMM3B42V6&V7doyouusethereal-timeproductorthe“highquality”product(whichlagsbyabout3months)? page9,line209:

Whydoyouusethe“standardizedmonthlymeanoftheprecipitation”?Standardizationshouldnotaffectyourcomputed correlationvalues

page9,line215:

“justifiesthechoiceofusingonlyregionsA1-4”

Idonotunderstandthereasoning.Seealsogeneralcommentabove

page10,line230:

“tohe”−>“tothe”

page17,line420:

“optimumvalue”−>“idealvalue”

SeetheoriginalKGEpublicationbyGuptaetal.(2009)howtheyusetheterms“optimal”and“ideal”

page20,line488,489:

“KGE’efficiencies”−>“KGE’values”

KGE=Kling-GuptaEfficiency

KGEefficiency=Kling-GuptaEfficiencyEfficiency(2x)

page22,line536:

“dominatedlimitedfactors”

Doyoumean“dominatinglimitingfactor”?Thishasadifferentmeaning

Ph.D HaraldKling HydrologistPöyryHydroConsulting,Hydropower,Kranichbergg.4Vienna,1120Austria