loading page

Propagation of CMORPH rainfall errors to REW streamflow simulation mismatch in the Upper Zambezi Basin
  • +2
  • Webster Gumindoga,
  • Tom Rientjes,
  • Alemseged Haile,
  • Paolo Reggiani,
  • Hodson Makurira
Webster Gumindoga
Universiteit Twente Faculteit Geo-Informatie Wetenschappen en Aardobservatie
Author Profile
Tom Rientjes
University of Twente
Author Profile
Alemseged Haile
International Water Management Institute
Author Profile
Paolo Reggiani
Universitat Siegen
Author Profile
Hodson Makurira
University of Zimbabwe
Author Profile

Abstract

This study investigates propagation effects of CMORPH rainfall estimation errors on streamflow simulation for a headwater catchment of the Zambezi River. Model simulations (2006-2012) by the Representative Elementary Watershed (REW) framework are carried out for uncorrected and for bias corrected CMORPH product (the Climate Prediction Center Morphing technique of the National Oceanic and Atmospheric Administration, or NOAA). As a benchmark to assessments, the model is run for in-situ observed rainfall obtained from 6 stations at a daily timestep. Analysis of CMORPH rainfall necessitates bias correction. A suite of performance indicators indicates that uncorrected CMORPH estimates show substantial augmentation of rainfall error to streamflow simulation mismatch whereas bias corrected estimates show attenuation of error. The ɛ-NSGAII algorithm is selected for single and multi-objective calibration to assess CMORPH error propagation to REW streamflow results. Improved hydrograph simulation is achieved by multi-objective calibration. Flow discharge simulation during the dry season shows more substantial error attenuation compared to wet season high flow discharge simulation. Further, this study shows that ratios of model based actual evapotranspiration over rainfall (ETa/R) and stream flow over rainfall (Qs/R) (runoff coefficient) at seasonal base change subject to selected uncorrected and corrected CMORPH. REW water storage (ΔS) is affected as well as calibrated model parameters. The paper provides new insights on propagation effects of satellite based rainfall errors in stream flow modelling.