Evaluation of Bias Correction Techniques for RCA4 Model of CORDEX-Africa Precipitation and Temperature Data in Case of Omo Gibe River Basin, Ethiopia

Abstract

The accurate assessment of climate change that impacts on water resources is fundamental to sustainable development. While Regional Climate Models (RCMs) are essential tools for this task due to their high resolution, their outputs contain significant biases that must be corrected. This is particularly critical in data-scarce regions like East Africa where the selection of optimal Bias Correction Methods (BCMs) remains largely unexplored and often relies on generalized recommendations. This study addresses this research gap by evaluating and identifying the most suitable BCM for the Omo Gibe River Basin in Ethiopia. The observed climate data (1990–2020) from 35 stations and precipitation and temperature variables from the Rossby Center regional Atmospheric Mode for African domain (RCA4) RCMs of the CORDEX-Africa project were utilized. The finding of the study demonstrated that the delta-change method outperformed other techniques, achieving exceptional performance metrics for bias-corrected historical data: for temperature Root-mean-square error (R²) is 0.95, Nash Sutcliffe efficiency (NSE) is 0.97, Coefficient of determination (RMSE) is 0.0028 and for precipitation (R² = 0.9, NSE = 0.95, RMSE = 0.0025). The study concluded that the delta-change method is the most robust approach for correcting both historical and future climate projections in the basin. Its application is therefore highly recommended for subsequent climate change impact studies on hydrology in the Omo Gibe Basin and similar regions in East Africa.

Keywords: RCM, bias correction; climate variables; Omo Gibe River Basin.