Comparison of bias correction methods to enhance CHIRP rainfall estimates for improved streamflow simulation at Ziway-Shalla catchment, Ethiopia

Ziway-Shalla Catchment, Rift Valley Basin, Ethiopia. Rainfall data availability significantly impacts the performance of rainfall-runoff models. Satellite rainfall estimates have the potential to fill data gaps if their systematic error is corrected. This study aims to compare the performance of five bias correction methods namely power transformation (PT), quantile mapping based on gamma distribution (QMG), daily translation (DAT), distribution transformation (DT), and linear scaling (LS) using Climate Hazards Group Infrared Precipitation (CHIRP) product. The effect of bias corrections on simulated streamflow was assessed using the Hydrologiska ByrånsVattenbalansavdelning (HBV) model. Results revealed the raw CHIRP contains large biases and the accuracy of bias-corrected rainfall data showed spatial variation across the study basin in Ethiopia. The QMG and PT methods outperformed other bias correction methods, whereas DT performed poorly in capturing the spatial pattern of annual average rainfall and its temporal variation. The streamflow simulated by the HBV model resulted in ∼11 % relative volume error (RVE) when the raw CHIRP product served as model input. Using bias-corrected CHIRP rainfall estimates improved the model performance in capturing the volume and pattern of the observed streamflow hydrograph and calibrated model parameter values changed when the rainfall input and bias correction methods varied. The results highlight the importance of bias correction methods for improved streamflow simulation. • Five bias correction methods were evaluated to improve CHIRP satellite rainfall estimates for more accurate streamflow simulations. • Raw CHIRP satellite rainfall data exhibited significant biases at multiple spatial and temporal scales over the Ziway-Shala Sub-basin. • Gamma distribution (QMG) and Power transformation (PT) bias correction methods effectively reduced the rainfall bias and resulted in a better performance of the HBV model for streamflow simulation. • The study highlights the impact of bias-corrected CHIRP data on improving both rainfall estimates and streamflow predictions.