| In the context of the SERVIR-Africa project, the SERVIR Arizona Team is developing streamflow forecast systems on African basins using Satellite Precipitation Products (SPP) to drive the models. These products have errors that need to be addressed before using them to drive hydrologic models.;An analysis of the errors of the Satellite Precipitation Products TMPA-3B42RT, CMORPH, and PERSIANN over Africa is presented, followed by bias correction and error reduction methods to improve the remote sensed estimates. The GPCP 1-degree-day reanalysis product was used as the rainfall truth dataset. The Bias Correction Spatial Downscaling (BCSD) method developed by Wood et al., was used successfully to reduce the errors of SPP.;The original and bias corrected estimates from the three SPP are used to calibrate and simulate three catchments of the Senegal River basin using HYMOD, finding that the use of bias corrected estimates produces a significant improvement in streamflow simulation. |
