RAINFALL-RUNOFF MODELING OF HAWASAA CATCHMENT

Abstract

TheSWATmodelishydro-dynamicandphysically-basedmodelforapplicationincomplexand largebasins.Modelinputsarea sfollows:rainfall,airtemperature,soilcbaracteristics, topography,vegetation,bydrogeologyandotherrelevantphysicalparametersand.TheSoiland WaterAssessmentTool2005(SWAT2005)wasdevelopedbytheAgriculturalResearchService o ftheUnitedStatesDepartmento fAgricultureanddistributedbytheUSEnvironmental ProtectionAgencyforwatershedmanagement.Themainobjectiveo fthisstudyistoestimate overallflowyieldo fHawasaacatchmentwhichislocatedintheMainEthiopianRiftvalleyin Ethiopia.Afterthedataqualitytest,meteorologicalandhydrologicaldataforcalibrationand validationwereusedtoestimatetheparametersintheSWATmodel.Theevaluationo fthe SWATmodelresponsetorainfall-runo frelationshipfromsensitiveanalysiswetakethemost tensensitiveflowparameterwhichaffecttherainfallrunoffrelationshipo fthecatchment. AmongthesensitiveflowparametersthesoilevaporationCompensationfactor(ESCO)were foundtob emoresensitiveparametertorunoff.WhileThresholdwaterdepthintheshallow aquiferforflow(GWQMN) ,BaseflowAlphaFactor(ALPHA-BF) ,thresholdwaterforwateri n theshallowaquiferforflow"revamp"percolationtodeepaquifertoaccrue(REVAPMN) ,Soil depthinmeter(SOLZ)andSCSrunoffcurvenumber(CN-2)themostinfluentialflow parametero fthecatchmentandotherflowparameterles sinfluencethanother.Themean monthlyandannualwateryieldsimulatedforabas eperiodfoundtobe25.8mmand309.0mm, respectively.Seasonalwateryieldsimulationresulted56.6mm,89.7mmand155.0mmfordry, intermediateandwetseason srespectively.Wateryieldisthetotalamounto fwaterleavingthe HRUandenteringmainchannelduringthetimestep.(WYLD )=SURQ+LATQ+GWQTLOSS.Fromourresult,thetotalamounto fwaterleavingtheHRUandenteringmainchannel duringthetimestep(WYLD )i s769.8mmH2O.Thepredictedhydrographwascalibrated againstobservedoneandthemodelparameterswereoptimizedforgoodsimulation.The calibrationandvalidationresultshowthatthecoefficiento fdeterminations(R^)andtheNashSutcliffesimulationefficiency(ENS)are0.76and0.72forcalibrationand0.77and0.73for validationthisshowsitisgoodsimulation.