water Article Flood Simulations and Uncertainty Analysis for the Pearl River Basin Using the Coupled Land Surface and Hydrological Model System YongnanZhu1,2,ZhaohuiLin2,3,YongZhao1,*,HaihongLi1,FanHe1,JiaqiZhai1, LizhenWang1andQingmingWang1 1 StateKeyLaboratoryofSimulationandRegulationofWaterCycleinRiverBasin,ChinaInstituteofWater ResourcesandHydropowerResearch,Beijing100038,China;[email protected](Y.Z.);[email protected](H.L.); [email protected](F.H.);[email protected](J.Z.);[email protected](L.W.);[email protected](Q.W.) 2 InternationalCenterforClimateandEnvironmentSciences(ICCES),InstituteofAtmosphericPhysics, ChineseAcademyofSciences,Beijing100029,China;[email protected] 3 CollaborativeInnovationCenteronForecastandEvaluationofMeteorologicalDisasters, NanjingUniversityofInformationScienceandTechnology,Nanjing210044,China * Correspondence:[email protected];Tel.:+86-10-6878-1370 AcademicEditors:GordonHuangandYuruiFan Received:23March2017;Accepted:28May2017;Published:1June2017 Abstract: The performances of hydrological simulations for the Pearl River Basin in China were analysedusingtheCoupledLandSurfaceandHydrologicalModelSystem(CLHMS).Threedatasets, includingEastAsia(EA),high-resolutiongaugesatellite-mergedChinaMergedPrecipitationAnalysis (CMPA)-Daily,andtheAsianPrecipitationHighly-ResolvedObservationalDataIntegrationTowards Evaluation(APHRODITE)dailyprecipitationwereusedtodrivetheCLHMSmodeltosimulatedaily hydrologicalprocessesfrom1998to2006.Theresultsindicatethattheprecipitationdatawasthemost importantsourceofuncertaintyinthehydrologicalsimulation. Thesimulatedstreamflowdrivenby theCMPA-Dailyagreedwellwithobservations,withaPearsoncorrelationcoefficient(PMC)greater than0.70andanindexofagreement(IOA)similaritycoefficientgreaterthan0.82atLiuzhou,Shijiao, andWuzhouStations. ComparisonoftheNash-Sutcliffeefficiencycoefficient(NSE)showsthatthe peakflowsimulationabilityofCLHMSdrivenwiththeCMPA-Dailyrainfallisrelativelysuperior tothatwiththeEAandAPHRODITEdatasets. Thesimulationresultsforthehigh-flowperiodsin 1998and2005indicatethattheCLHMSispromisingforitsfutureapplicationinthefloodsimulation andprediction. Keywords: coupled land surface-hydrology model; flood simulation; uncertainty analysis; PearlRiverBasin 1. Introduction Climatechangecausedbynaturalfactorsandhumanactivitieshasbeencontinuouslyaggravated in China [1,2], which has resulted in water shortages, drought and flood disasters, and other socio-economicproblems. Morefrequentandsevereextremehydrologicaleventshavesignificant impactsonsocialdevelopmentandhumanlivesandlivelihoods. Landsurfacehydrologicalmodels, importanttoolsinstudyingtheterrestrialwatercycleandtherelatedhydrologicalextremes,hasbeen widelyappliedforwatershedflooddisasterforecastingandwaterresourceprotection[3,4]. Abnormal precipitation is an important factor leading to flooding and drought. As the main prognosticvariableoftheclimatemodellingandprediction,precipitationistheprimarysourceof uncertaintyinlandsurfaceandhydrologicalsimulations[5–7]. Successfulsimulationofhydrological extremesunderachangingclimatedependsstronglyonspatial-temporalresolutionandtheaccuracy Water2017,9,391;doi:10.3390/w9060391 www.mdpi.com/journal/water Water2017,9,391 2of13 of precipitation information. China has recently developed a variety of gridded precipitation datasetswithhightemporal–spatialresolutionsbasedoninstrumentrecordsfromgroundstations and satellite products. These include the 1◦ × 1◦ gridded daily precipitation data established by Feng[8]with728stationsinChina;0.5◦ ×0.5◦ dailyprecipitationdataestablishedbyChen[9]using 753stationsinChina,and0.25◦ ×0.25◦ dailyprecipitationdatainChinapublishedbytheNational MeteorologicalInformationCentre[10]basedon2425stations. Thereareanothertwowidely-used precipitation datasets, one is the 0.5◦ × 0.5◦ daily precipitation data for East Asia established by Xie[11],basedonobservationaldatafromEastAsianstations,andanotheristhe0.25◦ ×0.25◦ Asian PrecipitationHighly-ResolvedObservationalDataIntegrationTowardsEvaluation(APHRODITE) datasetdevelopedjointlywiththeJapanMeteorologicalAgency[12]. These data have been widely applied for studying the water cycle, climate change, and related issues. However, significant uncertainties remain in the precipitation datasets from meteorologicalstationsusedforgeneratingthefinalrainfallproduct,whichwillhenceleadtohigh uncertaintiesinhydrologicalsimulations. ThispapertakesthePearlRiver—thethird-longestriver in China—as an example and applies the Coupled Land Surface and Hydrological Model System (CLHMS)tosystematicallyinvestigatetheimpactofdifferentprecipitationdatasetsonhydrological processsimulations. 2. ModelandMethod 2.1. StudyArea ThestudyregioninthispaperisthePearlRiverBasin,whichcontainsthethird-longestriver in Southern China (Figure 1). The Pearl River system includes the West, East, and North Rivers andthePearlRiverDelta. MeasuredfromtheheadwatersoftheWestRiver,thePearlRiversystem hasalengthofmorethan2300kmandabasinareaof447,000km2,withamildandrainyclimate, whichisaffectedbythesubtropicalmonsoonsystemandtropicalcyclones. Thebasinmeanannual precipitationreaches1525mm,andrainfallvariesthroughouttheyearwithmorethan80%ofannual totalprecipitationoccurringfromApriltoSeptember. Owingtothelargequantityofprecipitation intherainyseason,withhighstrengthandlongduration,floodsrapidlyconvergeinthemountains andhillsintheupperandmiddlereachesoftheriverwithonlyfewadjustableimpoundlakesinthe middlereaches. Thisconditiondirectlyendangersthedevelopedandpopulouscitiesandtownsinthe lowerreaches. SeveralseverefloodsoccurredalongthePearlRiverin1998,2000,2005,and2006that directlythreatenedhumanlifeandproperty. Accordingtoincompletestatistics,flooddisastersinthe PearlRiverthatoccurredbetween2000and2013causedmorethanRMB100billionindirecteconomic losses[13]. TheWestRiveristhewesterntributaryofthePearlRiver. LiuzhouandNanningStationsare locatedalongitstributaries,withcatchmentareasof45,413km2and72,656km2,respectively. Wuzhou Station,themostimportantcontrolsiteonthemainstreamoftheWestRiver,hasacatchmentareaof 327,006km2,accountingfor94.6%oftheriver’scatchmentarea. Theothertwomaintributariesofthe PearlRiveraretheEastandNorthRivers. BoluoStationislocatedonthelowerreachesoftheEast River,whichistheeasterntributary. ThecatchmentareaaboveBoluoStationis25,325km2,accounting forabout71.7%ofthetotalareaoftheEastRiverBasin. ThecatchmentareaaboveShijiaoStation, thecontrolstationforthelowerreachesofthenortherntributaryofthePearlRiver, is38,363km2 andaccountsfor82.1%ofthetotalareaoftheNorthRiverBasin. Thethreemainriversconvergein thePearlRiverDelta,wheretherivernetworkisstaggered,andthetributariesandSouthChinaSea simultaneouslyaffectthewaterflowandwaterlevel. Water2017,9,391 3of13 Water 2017, 9, 391 3 of 13 FFiigguurree 11.. TTooppooggrraapphhyy ooff tthhee PPeeaarrll RRiivveerr BBaassiinn aanndd llooccaattiioonn ooff tthhee mmaaiinn hhyyddrroollooggiiccaall ccoonnttrrooll ssttaattiioonnss.. 2.2. Model Description and Simulation Design 2.2. ModelDescriptionandSimulationDesign The Coupled Land Surface and Hydrology Model System (CLHMS) is applied in this study and The Coupled Land Surface and Hydrology Model System (CLHMS) is applied in this study was briefly introduced by Yang et al. [14], developed on the basis of the Land Surface Transfer Model and was briefly introduced by Yang et al. [14], developed on the basis of the Land Surface Transfer –Hydrologic Model System (LSX–HMS) model [15]. The CLHMS includes a large-scale LSX [16] and Model–HydrologicModelSystem(LSX–HMS)model[15].TheCLHMSincludesalarge-scaleLSX[16]and a fine-grid distributed HMS. The land surface model includes a two-layer vegetation module, a three- afine-griddistributedHMS.Thelandsurfacemodelincludesatwo-layervegetationmodule,athree-layer layer snow module, and a six-layer soil module. The LSX model calculates the surface energy balance snowmodule,andasix-layersoilmodule. TheLSXmodelcalculatesthesurfaceenergybalanceand and assigns the evaporation, runoff, infiltration, and soil moisture results to the HMS model. The assigns the evaporation, runoff, infiltration, and soil moisture results to the HMS model. The latter latter includes a terrestrial hydrologic module, a groundwater hydrologic module, and a channel- includesaterrestrialhydrologicmodule,agroundwaterhydrologicmodule,andachannel-groundwater groundwater interaction module. The terrestrial hydrologic module simulates overland flow and interactionmodule. Theterrestrialhydrologicmodulesimulatesoverlandflowandriverrunoff; the river runoff; the dynamic process of surface water flow is calculated using a two-dimensional dynamicprocessofsurfacewaterflowiscalculatedusingatwo-dimensionaldiffusionwaveineight diffusion wave in eight probable directions, and the channel flow velocity is described by the probabledirections,andthechannelflowvelocityisdescribedbytheManningequation. Groundwateris Manning equation. Groundwater is described as a single-layer aquifer and the combined water flux, describedasasingle-layeraquiferandthecombinedwaterflux,includingsurfacewater,isrepresented including surface water, is represented by Darcy’s Law. The spatial resolution of both the land surface by Darcy’s Law. The spatial resolution of both the land surface and hydrological models is set to 2a0nd× h2y0dkrmolotgoiacavlo midocdhealnsg iisn sgest ctaol e2s0. T× h2e0i nktmer atoc taiovnoibde tcwheaenngitnhge LscSaXleasn. dThHeM inSteisrabcatsioedn obnetpwreeednic ttehde LSX and HMS is based on predicted soil moisture and groundwater depth [15,17]. The CLHMS soilmoistureandgroundwaterdepth[15,17]. TheCLHMSaccuratelyreproducesnaturalhydrological accurately reproduces natural hydrological processes, water balance, and seasonal and inter-annual processes,waterbalance,andseasonalandinter-annualvariationinstreamflow. Ithasbeenverified variation in streamflow. It has been verified against historical data for the Yellow, Huai, Song-Liao, againsthistoricaldatafortheYellow,Huai,Song-Liao,andPearlRiverBasins[18–20]. and Pearl River Basins [18–20]. Parameters in the CLHMS model include surface elevation, soil texture, vegetation type, Parameters in the CLHMS model include surface elevation, soil texture, vegetation type, hydro- hydro-geologicalparameters,andotherlandsurfaceandhydrologicalcomponents. Thelandsurface geological parameters, and other land surface and hydrological components. The land surface soil in soilintheCLHMSmodelhassixlayersintheupper4.35m;thethicknessesfromthesurfacetothe the CLHMS model has six layers in the upper 4.35 m; the thicknesses from the surface to the lower lowerlayerare0.05,0.10,0.20,0.40,1.0,and2.5m. Thesoiltexturecharacterincludessixlayersof layer are 0.05, 0.10, 0.20, 0.40, 1.0, and 2.5 m. The soil texture character includes six layers of sand/silt/claycontentfromtheglobalsoilprofiledatabasegeneratedbyWebbetal.[21]. sand/silt/clay content from the global soil profile database generated by Webb et al. [21]. Thehydrologicalparameters,suchasslope,basinboundary,andelevationdeviationarederived The hydrological parameters, such as slope, basin boundary, and elevation deviation are derived fromtheUnitedStatesGeologicalSurvey(USGS)HYDRO1kdatabasewiththeZBalgorithm[22]. from the United States Geological Survey (USGS) HYDRO1k database with the ZB algorithm [22]. Otherphysicalparametersincludingporosity,Manningroughness,andhydraulicconductivityare Other physical parameters including porosity, Manning roughness, and hydraulic conductivity are calibrated on the basis of the newest version of the Harmonized World Soil Database (HWSD) calibrated on the basis of the newest version of the Harmonized World Soil Database (HWSD) developedbytheFoodandAgriculturalOrganization(FAO)oftheUnitedNationsandInternational developed by the Food and Agricultural Organization (FAO) of the United Nations and International Institute for Applied System Analysis (IIASA). Launched in partnership with the Institute of Soil Institute for Applied System Analysis (IIASA). Launched in partnership with the Institute of Soil Science,ChineseAcademyofSciences,HWSDprovidesthemostrecent1:1,000,000-scalesubsoiland Science, Chinese Academy of Sciences, HWSD provides the most recent 1:1,000,000-scale subsoil and topsoilmapofChina[23,24]. topsoil map of China [23,24]. Three high-resolution precipitation datasets—the East Asia (EA), APHRODITE, and CMPA- Daily precipitation data—were selected to investigate the impact of precipitation on the hydrological Water2017,9,391 4of13 Threehigh-resolutionprecipitationdatasets—theEastAsia(EA),APHRODITE,andCMPA-Daily precipitationdata—wereselectedtoinvestigatetheimpactofprecipitationonthehydrologicalprocess simulationusinglandsurface–hydrologicalcoupledmodels. Abriefdescriptionoftherainfalland othermeteorologicalforcingdatasetsisgivenbelow. (1) EastAsiaDailyPrecipitationData The East Asia precipitation data was provided by the National Oceanic and Atmospheric Administration(NOAA)ClimatePredictionCenter(CPC)[11]. Ithasaspatialresolutionof0.5◦ ×0.5◦ andspatialextentof5◦N–60◦N,65◦E–155◦Eandcoverstheperiod1962–2006.Thedatawerecollected byanalysingtheoptimaldifferencevalueswiththeobservedprecipitationdatafrom2200stationsin EastAsia,including730meteorologicalstationsfromtheChinaMeteorologicalAdministration(CMA) and1000hydrologicalobservationstationsfromtheYellowRiverConservancyCommissionofthe MinistryofWaterResourcesofthePeople’sRepublicofChina. (2) APHRODITEPrecipitationDataSet The APHRODITE plan was implemented jointly by the Research Institute of Humanity and Nature(RIHN)andtheMeteorologicalResearchInstituteofJapanMeteorologicalAgency(MRI/JMA). TheseagenciesestablishedadatasetdescribingtheprecipitationfeaturesoftheAsianmonsoonregion (MA), Middle Asia (ME), Russia (RU), and Japan (JR) by integrating the observational data in the Asianregionsandprecipitationstationsfromdifferentcountries. Here,theAPHRODITEdatasetfor monsoonAsia(APHRO-MA)wasselected,whichhasaspatialresolutionof0.25◦ ×0.25◦;aspatial extentof15◦ S–55◦ N,60◦ E–150◦ E;anddurationof1951–2007. Thisdatasetiscomposedprimarilyof stationdataprovidedbytheCMAandGlobalTelecommunicationSystem(GTS)datapreparedbythe WorldMeteorologicalOrganization[12]. (3) High-ResolutionGauge–SatelliteMergedCMPA-DailyData TheCMPA-Dailyprecipitationdatasethasaspatialresolutionof0.25◦ ×0.25◦,aspatialextentof 0◦ N–60◦ N,60◦ E–160◦ E,anddurationof1998–2013. Itusesdatafrom2425national-levelground weatherstationsandtheCPCMorphingTechnique(CMORPH)datadevelopedbytheNOAAGlobal Precipitation Climatology Project (GPCP). The errors in CMORPH satellite precipitation data are correctedusingtheprobabilityanddensitymatchingmethodandaregeneratedwiththeoptimal differencemethodonthebasisoftheclimatebackgroundfield[9]. To compare the uncertainty from precipitation data in land surface hydrological simulations, ThreesetsofcontrastingnumericalexperimentsbasedonCLHMShavebeendesignedtosimulate thedailywatercycleprocessforthePearlRiverduringanine-yearperiodof1998–2006. Thethree setsofexperimentsusedtheEA,APHRO-MA,andCMPA-Dailydataasvaryingprecipitationfields, whereastheothermeteorology-drivendataremainedthesame,asshowninTable1. Foursimulation experimentsweredesignedtocomparetheinfluenceofdifferentprecipitationdatasetsandmodel resolutiononwatercycleprocesssimulations. Intheexperiments,allofthemeteorologicaldatawere interpolatedtothemodelresolution;thetimestepsforthehydrologicalandlandsurfacemodelswere 24hand15min,respectively. Table1.Introductiontotheexperimentdesign. No. TestName Resolution PrecipitationDataSources MeteorologicalParameters 1 EA 20km×20km Xieetal.[11] 2 APHRO-MA 20km×20km Yatagaietal.[12] CN05dailytemperaturedataset[25]; 3 CMPA-Daily(20km) 20km×20km 6-hNCEP-NCARreanalysisdata[26] Shenetal.[9] 4 CMPA-Daily(10km) 10km×10km ThetemperaturedatawerebasedonthedailytemperaturedatasetoverChina(CN05)[25]from the CMA. The sub-daily data for temperature was disaggregated using statistical downscaling of Water2017,9,391 5of13 the global 3-h temperature dataset from Princeton University [27]. The near surface wind speed, humidity,airpressure,radiationflux,andotherbasicmeteorologicalforcingdatawereobtainedfrom the6-hNationalCentresforEnvironmentalPrediction–NationalCentreforAtmosphericResearch (NCEP–NCAR)reanalysisdata[26]. The full model was integrated with three separate processes to obtain near equilibrium groundwatertables. Thefirstphaseisthe“coldstartprocess”,inwhicha50-yearrunwasperformed withthefullmodelandobservedmeteorologicaldatafor1998–2006,forthemodeltoreachinitial surfacewaterbalance. Thesecondisthe“spin-upprocess”,inwhichonlythegroundwatermodel wasspunupfor5000yearsusingthecoldstartresultastheforcing,inorderforthemodeltoreach equilibriumforgroundwaterbalance. Usingthemodeloutputfromthe“spin-upprocess”asinitial conditions,theCLHMSwasthenrunfrom1998to2006fortwocycles,for18yearsintotal,andthe observeddailyprecipitationdata,CN05dailytemperaturedata,andNCEP/NCARsix-hourreanalysis datawereusedasforcingdata. Toreducethemodelbiascausedbymodelinitialization, onlythe modelsimulationresultsfromthesecondcycle(year10toyear18)wereusedtoanalysethewater cyclefrom1998to2006. To correctly compare the influence of different precipitation forcings on the simulation performanceofthelandsurface–hydrologicalmodel,thewaterbalanceindex(WBI),Nash–Sutcliffe efficiencycoefficient(NSE),Pearsonproduct-momentcorrelationcoefficient(PMC),IOAsimilarity coefficient,andnormalizedrootmeansquareerror(NRSE)wereusedinthisstudy. Amongthese, theWBIprimarilyreflectstheabilityofthemodeltosimulatethewaterquantitybalanceprocess;the NSEreflectsthesimulationabilityforpeakflow; thePMCandIOArepresentthetimecorrelation andsimilaritybetweentheobservedandsimulatedstreamflow, respectively; andtheNRSEisthe rootmeansquareerroradjustedbytheaveragevalueoftheobservationsequenceateachsite,which facilitatesthecomparisonofdifferentsites. Foreachofthesefiveindices,valuesclosertooneindicate highercapabilitiesofthemodelsimulation. Theformulaeforthefiveindicesareasfollows: ∑N P WBI= i=1 i; (1) ∑N O i=1 i ∑N (P −O)2 NSE=1.0− i=1 i i ; (2) ∑N (cid:0)O −O(cid:1)2 i=1 i ∑N (P −P)(cid:0)O −O(cid:1) PMC= i=1 i i ; (3) (cid:104)∑N (P −P)2(cid:105)0.5(cid:104)∑N (O −O)2(cid:105)0.5 i=1 i i=1 i ∑N (P −O)2 IOA=1.0− i=1 i i ; (4) ∑iN=1(cid:0)(cid:12)(cid:12)Oi−O(cid:12)(cid:12)+(cid:12)(cid:12)Pi−O(cid:12)(cid:12)(cid:1)2 (cid:118) NRSE=1.0−(cid:117)(cid:117)(cid:116) 1 ∑N (cid:18)Pi−Oi(cid:19)2 (5) N O i=1 among them, P and O are the values at the ith day simulated and observed daily streamflow, i i respectively;PandOaretheaveragevaluesofthesimulatedandobservedsequences;Nisthetotal numberofsamples. Inthisstudy,weselectedLiuzhou,Nanning,andWuzhouStationsontheWest River,BoluoStationontheEastRiver,andShijiaoStationontheNorthRiverasthemaincontrolsites forthePearlRiverBasin. TherivernetworkinthePearlRiverDelta,wheretheWest,East,andNorth Riversconverge,isstaggeredandisinfluencedbytheseawaterintrusion,sothehydrologicalstation inthePearlRiverDeltawasnotselectedforthisstudy. Water2017,9,391 6of13 3. AnalysisofSimulationResults 3.1. EvaluationofDailySimulationPerformance Figure 2 shows the simulated daily streamflow under three different precipitation forcings comparedwiththeobserveddailystreamflowatthemainstationsinthePearlRiverBasinduring 1998–2006. Table2presentstheperformanceoftheCLHMSmodelbasedontheWBI,NSE,PMC,IOA sWimateirl a2r0i1t7y,, 9a, n39d1 NRSEforNanning,Liuzhou,Wuzhou,Shijiao,andBoluoStations. 6 of 13 FFiigguurree 22.. SSccaattteterprplolotst soof fa acocmompapraisroisno nbebtweteweene ntheth oebosebrsveerdv esdtresatrmeaflmowflo awnda nsdimsuimlatuiolant iofrnomfr othme tChoeuCploeudp LleadndL SaunrdfacSeu arnfadc eHyadnrdolHogyidcarol Mlogoidceall SMysotedmel (CSyLsHteMmS)( mCLodHeMl fSo)r (ma)o Ldieulzhfooru;( (ab)) LWiuuzzhhoouu;; ((bc)) WShuijziahoo;u a;n(dc) (Sdh) iBjiaoolu;oa nSdta(tdio)nBso oluno thSeta Ptieoanrls Ronivtehr efrPoemar 1l9R9i8v etor f2r0o0m6 u19n9d8erto d2if0f0e6reunnt dperrecdipififtearteionnt pforercciipngitsa.t ionforcings. ATallbtlhe r2e. eEvtealsutastisohno owf ethde gpoerofdormsiamnucel aotfi othne aCboiuliptlyedin Latnedrm Susrofafcteh aendW HById.rFoolorgtihcael Msimoduella Styiostnemu sing theC(MCLPHAM-DSa).i lyrainfallastheforcing,thePMCcorrelationcoefficientsfortheNanning,Liuzhou, andBoluoStationswere0.63,0.70,and0.65,respectively,whicharethehighestamongthethreesets Hydrologic Station Nanning Liuzhou Wuzhou Shijiao Boluo ofsimulationswithdifferentEpAr ecipitatio0n.91fo rcings0..9In6 contra0s.9t,8t hePM0.C99c orre0la.8t3io ncoefficientsat WuzhouandShijiWaoBSI tatiAoPnHsfRoOr-tMheAC MPA0.9-D3 ailywe1r.0e10 .92and1.000.8 2,resp0.e9c9t ively0,.w86h ichshowsbetter simulationabilityinthemaCinMsPtAre ams. 0.94 1.00 1.1 1.13 0.84 EA 0.36 0.44 0.78 0.61 0.27 NSE APHRO-MA 0.36 0.42 0.76 0.62 0.27 CMPA 0.38 0.47 0.81 0.67 0.25 EA 0.62 0.68 0.89 0.78 0.63 PMC APHRO-MA 0.61 0.66 0.89 0.79 0.64 CMPA 0.63 0.70 0.92 0.82 0.65 EA 0.75 0.80 0.92 0.86 0.76 IOA APHRO-Ma 0.75 0.78 0.92 0.86 0.78 CMPA 0.75 0.82 0.95 0.90 0.77 EA −0.06 −0.21 0.49 0.20 0.14 NRSE APHRO-MA −0.06 −0.23 0.47 0.21 0.14 CMPA −0.06 −0.17 0.53 0.20 0.10 Water2017,9,391 7of13 Table 2. Evaluation of the performance of the Coupled Land Surface and Hydrological Model System(CLHMS). HydrologicStation Nanning Liuzhou Wuzhou Shijiao Boluo EA 0.91 0.96 0.98 0.99 0.83 WBI APHRO-MA 0.93 1.01 1.00 0.99 0.86 CMPA 0.94 1.00 1.1 1.13 0.84 EA 0.36 0.44 0.78 0.61 0.27 NSE APHRO-MA 0.36 0.42 0.76 0.62 0.27 CMPA 0.38 0.47 0.81 0.67 0.25 EA 0.62 0.68 0.89 0.78 0.63 PMC APHRO-MA 0.61 0.66 0.89 0.79 0.64 CMPA 0.63 0.70 0.92 0.82 0.65 EA 0.75 0.80 0.92 0.86 0.76 IOA APHRO-Ma 0.75 0.78 0.92 0.86 0.78 CMPA 0.75 0.82 0.95 0.90 0.77 EA −0.06 −0.21 0.49 0.20 0.14 NRSE APHRO-MA −0.06 −0.23 0.47 0.21 0.14 CMPA −0.06 −0.17 0.53 0.20 0.10 TheNSEefficiencycoefficientshowsthatthepeakflowsimulationabilityoftheCMPA-Dailytest isobviouslysuperiortothatoftheEAandAPHRO-MAdatasets. However,itsabilitytosimulatethe peakflowinmainstreamsandtributariesshowsacleardifference. AccordingtotheCMPA-Daily precipitation forcing, the simulated peak flow was larger than the EA- and APHRO-MA-driven simulatedflowfor1998,2000,2001,2004,and2005. Takingthe1998catastrophicfloodintheWest Riverasanexample,WuzhouStationonthemainstreamoftheWestRiverobservedthefirstpeak on28Juneataflowof51,500m3/s; thearrivaltimeofthepeaksimulatedintheEAprecipitation forcing test was 25 June, which is three days prior to observation. The simulated peak flow was 27,933 m3/s, which is 45.7% smaller than the observation; the peak arrival time simulated in the APHRO-MAforcingtestwasthesameasthatintheEAtest. Thepeakflowwas26,423m3/s,whichis 48.6%smallerthantheobservation. ThearrivaltimeofthepeaksimulatedintheCMPA-Dailytestwas 21June,whichisonedayaheadoftheobservation. Thesimulationpeakflowwas46,136m3/s,which is10.4%smallerthantheobservation. ThisfloodalsooccurredintheNorthRiverduringthesame year. ShijiaoStationobservedthepeakarrivaltimeon26Junewithapeakflowof12,200m3/s;the simulatedpeakarrivaltimeintheEArainfall-driventestwas25June,whichisonedayaheadofthe observation. Thesimulationpeakflowwas6681m3/s,whichis45.2%smallerthantheobservation; thepeakarrivaltimesimulatedinAPHRO-MAtestwasthesameasthatintheEAtest. Thepeakflow was7621m3/s,whichis37.5%smallerthantheobservation. InthesimulationfromtheCMPA-Daily test,thepeakarrivaltimewas24June,andthepeakflowwas9393m3/s,whichis23.0%smallerthan theobservation. Acomparisonoftheofflowsimulationresultsimilaritiesunderdifferentprecipitationdrivers withtheobservationflowsequencerevealedthattheIOAsimilarityindicatorsforthethreetestsat BoluoStationwerenear0.76.ThosefortheEAandAPHRO-MAtestsatNanning,Wuzhou,andShijiao Stationswere0.75,0.92,and0.86,respectively. Thevalueswere0.75,0.95,and0.90,respectively,forthe CMPAtest. ComparingthesimulationresultsfromLiuzhouStation,thesimilarityfortheAPHRO-MA testwas0.78,whichwasthelowestofthethreetests. TheIOAsimilarityfortheCMPAtest,0.82,was thehighest. TheIOAindicatorsforallstationsunderdifferentprecipitationdriverswerehigherthan 0.75,indicatingthatallsimulationresultsandobservationswerehighlysimilar. TheCMPAsimulation resultsweresuperiorthoseofEAandAPHRO-MA. Toanalysethecausesforthedifferencesinthesimulatedstreamflowresults,wecomparedthe differencesinmonthlyanddailyprecipitationdistributionfromtheEA,APHRO-MA,andCMPA-Daily datasets. Taking the rainstorm period in June 1998 as an example, as shown in Figure 3a–c, the maximumrainfallareaappearedatthenorthernbranchoftheWestRiver. However,somedifferences Water2017,9,391 8of13 inspatialdistributionofthethreeprecipitationdatasetswerenoted. Moreover,theprecipitationgrid datashowedadivergenceintherainfallamount. ThemaximummonthlyrainfallofEA,APHRO-MA, and CMPA-Daily datasets were 754.59, 681.22, and 967.43 mm, respectively. Figure 3d shows the variationintheregionalaveragedailyprecipitationinJune1998. ThedailyrainfalleventsofEAwere essentiallysimilartothoseofAPHRO-MA,exceptforarainstormon16–26June. Comparedwiththe othertwodatasets,theCMPA-Dailyrainfalleventsshowedgreaterdifferencesatthebeginningof themonth. For1998–2006,theCMPA-Dailyaverageannualprecipitationintheriverwas1759mm; the values from the EA and APHRO-MA data were 1498 mm and 1525 mm in the same period, respectively. TheEAandAPHRO-MAprecipitationdatavalueswere14.8%and13.3%smallerthan theCMPA-Dailydata,respectively. The1998–2006summerprecipitationwasalsoclearlydifferent, withtheCMPA-Daily,EA,andAPHRO-MAprecipitationdatashowingaveragesummerprecipitation (JuWneat–erA 2u01g7,u 9s, t3)91o f9.64,8.12,and8.13mm/d,respectively. 8 of 13 FigFuigreur3e. 3C. oCmompapraisriosnono fofp prerecicpipitiatatitoionni inn JJuunnee 11999988 ffoorr mmoonntthhllyy rraaininfafalll lddisitsrtirbiubtuiotino nfrofrmom (a)( aE)aEsta AstsAias ia (EA(E)A;()b; )(bA) sAiasnianP rPerceipciiptaittaitoinonH Higighhlyly-R-Reessoollvveedd OObbsseerrvvaattiioonnaal lDDaatata InIntetgegrartaiotino nToTwowaradrsd EsvEavluaalutiaotnio n dadtaasteatsefto rfoMr MonosnosoononA Asisaia( A(APPHHRROO--MMAA));; ((cc)) hhiigghh--rreessoolluuttiioonn ggaauuggee ssaatetellliltiet-em-meregregde dChCihnain MaMergeerdge d PrePcriepciitpaittiaotnionA Ananlaylsyissis( C(CMMPPAA)-)D-Daaililyy;; aanndd ((dd)) tthhee rreeggiioonnaall ddaailiyly aavveerargaeg eprperceicpiiptaittiaotnio nforf otrhrteher ee prepcriepciitpaittiaotniofno rfocirncignsg.s. 3.2. Evaluation of Hydrological Extreme Simulation Performance 3.2. EvaluationofHydrologicalExtremeSimulationPerformance To evaluate the CLHMS simulations of flood processes on the Pearl River, this study used the ToevaluatetheCLHMSsimulationsoffloodprocessesonthePearlRiver,thisstudyusedthe 20 × 20 km and 10 × 10 km grid resolution CLHMS models and the CMPA-Daily precipitation data to 20×20kmand10×10kmgridresolutionCLHMSmodelsandtheCMPA-Dailyprecipitationdatato simulate the streamflow on the Pearl River for 11 June to 1 August 1998 and 20 May to 20 July 2005. simulatethestreamflowonthePearlRiverfor11Juneto1August1998and20Mayto20July2005. In June 1998, a flood with a return period of 100 years occurred on the West and North River InJune1998, afloodwithareturnperiodof100yearsoccurredontheWestandNorthRiver tributaries of the Pearl River. The regional rainfall quantity exceeded 400 mm in the Liujiang and tributaries of the Pearl River. The regional rainfall quantity exceeded 400 mm in the Liujiang and Guijiang Rivers, at the rainstorm centre. At Wuzhou Station, the maximum observation flow was GuijiangRivers, attherainstormcentre. AtWuzhouStation, themaximumobservationflowwas 52,900 m3/s with a recurrence interval of 100 years. 52,900m3/switharecurrenceintervalof100years. From 9 to 25 June 2005, catastrophic flooding occurred on the Pearl River owing to continuous From9to25June2005,catastrophicfloodingoccurredonthePearlRiverowingtocontinuous rainstorms. At Longmen Station on the East River, the maximum precipitation reached 1442 mm. The raipnrsetcoirpmitas.tioAnt aLt otnhge mraeinnsStotramtio cnenotnret hone Ethaes tWReisvte Rr,ivthere wmaasx 4im00u–m500p mremcip. iAtat tWionuzrheoauc hSetdati1o4n4, 2thme m. Thpeeparke cfilpowita trieoanchaetdt h5e3,r9a0i0n smto3r/sm, wcehnictrhe roannkths eaWs ethste Rsievceorndw-hasig4h0e0st– 5f0lo0wm smin.cAe tthWe usztahtoiounS wtaatiso n, theesptaebalkisflheodw. Trehaec lhaerdge5st3 ,f9lo0o0dm d3u/rsin,gw thhiec hparastn 2k0s yaesatrhse oscecucorrnedd- hoing hthees tEflaostw Risvinerc,e atfhfeectsitnagt io30n.3w2 as estmabillliisohne dp.eTophlee laanrgde csatuflsoiondg ddiurreicntg ecthoenopmasitc 2lo0syseeas rosfo RcMcuBr r3e1d.4o5n btihlleioEna [s1t3R].i v er,affecting30.32million peopleFaingdurcea 4u ssihnogwdsi trheec tseimcounlaotmedic alnods soebssoerfvReMd rBai3n1s.t4o5rmbi lpliroonce[s1s3e]s. at the main stations of the Pearl River from 11 June to 1 August 1998. Table 3 presents the simulation performance based on WBI, NSE, PMC, IOA similarity, and NRSE during the same period. The rainfall was concentrated mainly on the West and North Rivers; flooding did not occur on the East River. CLHMS simulated the flood process better at Liuzhou, Wuzhou, and Shijiao Stations, and the simulation results at Boluo Station on the East River were also consistent with the observations. During the rainstorm, three peak flooding events occurred continuously at Liuzhou Station on the West River. The third event, which occurred on 24 June, had the largest peak of the flooding process with a streamflow of 20,000 m3/s. The simulation results with 20 km resolution showed that the highest peak at Liuzhou Station occurred during the second flood event on 21 June, with a streamflow of 14,838 m3/s. In addition, the third flood event, occurring on 23 and 24 June, was lower than the previous event; this result contradicts the observations. In contrast, the simulation result from the 10 km grid resolution showed stronger similarity to the observed flood events at Liuzhou Station. The largest peak occurred on 23 Water2017,9,391 9of13 Figure4showsthesimulatedandobservedrainstormprocessesatthemainstationsofthePearl Riverfrom11Juneto1August1998. Table3presentsthesimulationperformancebasedonWBI,NSE, PMC,IOAsimilarity,andNRSEduringthesameperiod. Therainfallwasconcentratedmainlyon theWestandNorthRivers; floodingdidnotoccurontheEastRiver. CLHMSsimulatedtheflood processbetteratLiuzhou,Wuzhou,andShijiaoStations,andthesimulationresultsatBoluoStationon theEastRiverwerealsoconsistentwiththeobservations. Duringtherainstorm,threepeakflooding eventsoccurredcontinuouslyatLiuzhouStationontheWestRiver. Thethirdevent,whichoccurredon 24June,hadthelargestpeakofthefloodingprocesswithastreamflowof20,000m3/s. Thesimulation resultswith20kmresolutionshowedthatthehighestpeakatLiuzhouStationoccurredduringthe secondfloodeventon21June,withastreamflowof14,838m3/s. Inaddition,thethirdfloodevent, occurringon23and24June,waslowerthanthepreviousevent;thisresultcontradictstheobservations. Water 2017, 9, 391 9 of 13 Incontrast,thesimulationresultfromthe10kmgridresolutionshowedstrongersimilaritytothe observJuende,fl wohoidche vise onntes datayL piurziohro tuo tShtea otibosne.rvTehde elvaerngte. sTthpee paekako cflcouwrr wedaso 1n6,26325J umn3e/s,, wanhdic thhei ssiomnueladtaioynp rior totheeorrbosre wrvaes d16e.9v%en. t. Thepeakflowwas16,625m3/s,andthesimulationerrorwas16.9%. Figure 4. Flood simulation results for (a) Liuzhou; (b) Wuzhou; (c) Shijiao; and (d) Boluo Stations on Figure4.Floodsimulationresultsfor(a)Liuzhou;(b)Wuzhou;(c)Shijiao;and(d)BoluoStationson the Pearl River in 1998. thePearlRiverin1998. Table 3. Evaluation of hydrological extremes simulation in 1998. Table3.Evaluationofhydrologicalextremessimulationin1998. Hydrologic Station Liuzhou Wuzhou Shijiao Boluo 20 km 0.70 1.02 0.87 0.78 HydrologicWSBtaIt ion Liuzhou Wuzhou Shijiao Boluo 10 km 0.82 1.00 0.90 0.87 20km 20 km 0.700.62 0.19.30 2 0.80 0.870.22 0.78 WBI NSE 10km 10 km 0.802.76 0.19.00 0 0.95 0.900.60 0.87 20km 20 km 0.602.90 0.09.89 3 0.91 0.800.88 0.22 NSE PMC 10km 10 km 0.706.90 0.09.59 0 0.98 0.950.88 0.60 20 km 0.77 0.96 0.90 0.74 IOA20 km 0.90 0.98 0.91 0.88 PMC 10km 10 km 0.900.90 0.09.59 5 0.98 0.980.84 0.88 20 km 0.52 0.96 0.62 0.67 NRS2E0 km 0.77 0.96 0.90 0.74 IOA 10 km 0.62 0.95 0.81 0.76 10km 0.90 0.95 0.98 0.84 20km 0.52 0.96 0.62 0.67 The meNasRuSrEed peak flow at Wuzhou Station occurred on 28 June, when the peak flow was 51,500 10km 0.62 0.95 0.81 0.76 m3/s. The peak occurred on 27 June for both the 10 km and 20 km grid resolution simulations one day prior to the observed event. The simulated peak flow with 10 km resolution was 50,988 m3/s, and the simulation error was only −0.4%. The simulation error with 20 km resolution was −10.4%. During the flood events in 1998, the peak flow at Shijiao Station was 12,200 m3/s, and the peak arrival date was 26 June. The 20 km resolution simulation results showed the peak arrival date to be 24 June, which is two days prior to the observed event. The peak flow was 9393 m3/s, which is 23.0% less than the observed flow. The peak flow was 12,259 m3/s with 10 km resolution simulation, which is 0.5% larger than the observation, and the peak arrival time was consistent with the observation. The 20 km grid resolution simulation error in peak flow was larger. Generally, the 10 km resolution simulation of flooding of the Pearl River in 1998 is more accurate. Figure 5 and Table 4 present the results for the flood season from 20 May to 20 July 2005. We compared the observed and simulated streamflow from the coupled land surface–hydrological Water2017,9,391 10of13 The measured peak flow at Wuzhou Station occurred on 28 June, when the peak flow was 51,500m3/s. Thepeakoccurredon27Juneforboththe10kmand20kmgridresolutionsimulations onedaypriortotheobservedevent. Thesimulatedpeakflowwith10kmresolutionwas50,988m3/s, andthesimulationerrorwasonly−0.4%. Thesimulationerrorwith20kmresolutionwas−10.4%. Duringthefloodeventsin1998,thepeakflowatShijiaoStationwas12,200m3/s,andthepeakarrival datewas26June. The20kmresolutionsimulationresultsshowedthepeakarrivaldatetobe24June, whichistwodayspriortotheobservedevent. Thepeakflowwas9393m3/s,whichis23.0%lessthan theobservedflow. Thepeakflowwas12,259m3/swith10kmresolutionsimulation,whichis0.5% largerthantheobservation,andthepeakarrivaltimewasconsistentwiththeobservation. The20km gridresolutionsimulationerrorinpeakflowwaslarger. Generally,the10kmresolutionsimulationof floodingofthePearlRiverin1998ismoreaccurate. Figure 5 and Table 4 present the results for the flood season from 20 May to 20 July 2005. Wecomparedtheobservedandsimulatedstreamflowfromthecoupledlandsurface–hydrological model at different grid resolutions. The results showed that 10 km grid resolution provides more accurateWpateera 2k017fl, o9, w391a ndtimingsimulationsandthatofthesimulatedpeakflow,whichwa1s0 olfo 1w3 erthan theobservationwhenusingthe20kmresolutiongrid. Thesimulationsusingthe20kmresolution model at different grid resolutions. The results showed that 10 km grid resolution provides more showedacacupreaatek pfleoawk flaotwW aundz htiomuinSgt astiimounlaotifo3n5s ,0an2d5 tmha3t/ osf, twheh iscimhuisla3te4d. 3p%easkm floawlle, rwthhiachn wthaes lloawrgeer stpeak, 53,300mth3a/n st,hme eoabssuerrveadtioonn w2h2eJnu nuesi.nWg thheen 20u skimng rtehsoelu1t0ioknm grride.s oTlhuet isoimnuglraitdio,ntsh uessinimg uthlea t2e0d kpme akflow atWuzhreosuoluSttiaotni oshnowweads a5 6p,e1a2k0 flmow3/ ast, Wwuhzihcohu iSst5at.i3o%n olfa r3g5,e0r25t hma3n/s,t hwehiocbh sies r3v4e.3d%e svmeanllte.rF toharnb tohteh 10km largest peak, 53,300 m3/s, measured on 22 June. When using the 10 km resolution grid, the simulated and20kmgridresolutions,thesimulatedarrivaldateofthelargestpeakflowatWuzhouStationwas peak flow at Wuzhou Station was 56,120 m3/s, which is 5.3% larger than the observed event. For both 20June,whichistwodayspriortoobservation. Theobservedpeakarrivedon24JuneatBoluoStation, 10 km and 20 km grid resolutions, the simulated arrival date of the largest peak flow at Wuzhou andthepeakflowwas7760m3/s. Whenusing20kmand10kmresolutiongrids,thesimulatedpeak Station was 20 June, which is two days prior to observation. The observed peak arrived on 24 June at arrivaldBoaltueos Swtaetiroen2, a1nJdu tnhee apneadk 2fl3owJu wnaes, r7e7s6p0 emc3t/isv. eWlyh.enT huseinsgim 20u klamte adndp 1e0a kkmfl orewsowlutaiosn1 g2r,i4d3s0, tmhe3 /sand 10,391msim3/usl,atoevd epreeaskt iamrraivtaeld dbatyes6 w0%erea 2n1d Ju3n3e. 9a%nd, 2re3 sJpunece,t irveseplye.ctTivheley.s Timheu sliamtiuolanterde spuealtks fwlower ewgase nerally higher t1h2,a4n30 tmhe3/so abnsde r1v0,a3t9i1o nms3./s,A ovcecroersdtiimnagtetdo btyh 6e0%C hainnda 3w3.9a%te, rrersepseoctuivrecleys. Tbhuel lseimtiunl,at2io5nm reesdulitus m- and were generally higher than the observations. According to the China water resources bulletin, 25 large-sizedreservoirswereestablishedonthePearlRiverfrom2001to2005,includingfivelarge-scale medium- and large-sized reservoirs were established on the Pearl River from 2001 to 2005, including reservoirs[13]. Therefore,theregulationandcontrolonthepeakarrivaltimeandpeakflowaremore five large-scale reservoirs [13]. Therefore, the regulation and control on the peak arrival time and importantindeterminingthestreamflow. peak flow are more important in determining the streamflow. Figure 5. Flood simulation results at (a) Liuzhou; (b) Wuzhou; (c) Shijiao; (d) and Boluo Stations in Figure5.Floodsimulationresultsat(a)Liuzhou;(b)Wuzhou;(c)Shijiao;(d)andBoluoStationsinthe the Pearl River in 2005. PearlRiverin2005. Table 4. Evaluation of hydrological extreme simulation in 2005. Hydrologic Station Liuzhou Wuzhou Shijiao Boluo 20 km 0.77 1.01 0.96 1.09 WBI 10 km 0.66 1.09 0.88 1.04 20 km −0.08 0.61 0.74 −0.67 NSE 10 km −0.46 0.66 0.61 0.40 20 km 0.43 0.83 0.86 0.46 PMC 10 km 0.20 0.84 0.81 0.85 20 km 0.33 0.75 0.88 0.54 IOA 10 km 0.08 0.88 0.83 0.87 20 km 0.21 0.58 0.59 −0.31 NRSE 10 km 0.08 0.60 0.49 0.22