Evaluating the Effect of Bias Correction Methods in Regional Climate Model Outputs for Rainfall- Runoff Modeling In Katar Watershed Ethiopia

Abstract Identifying a suitable bias correction method is important to provide reliable inputs for evaluation of climate change impact. This study evaluates five precipitation bias correction methods (BCM) and three temperature BCM for Katar catchment. The BCMs were evaluated using several statistical measures such as, bias (P BIAS ), root mean squared error (RMSE), mean absolute error (MAE), coefficient of variation (CV), personal correlation coefficient (R 2 ), and relative volume error (RVE). The annual rainfall bias of the models varies between 7.5% and 257.93% suggesting overestimation. The result showed that the methods used to correct bias improve the RCM-simulated rainfall and temperature to a certain degree in terms of frequency and time series based statics. The raw RCM- simulated precipitation overestimates the wet month observed precipitation in all RCM- outputs. All bias correction methods reduce the overestimation of the raw RCM-simulated precipitation in all models. From the result ECDF performed better than the other bias correction methods based on the corrected annual, monthly, and daily rainfall comparison followed by PT, DM, LOCI, and LS. The temperature bias-correction method's effect was insignificant. The performance of raw RCM precipitation and temperature for stream flow was very poor with a value of NSE (0.2), R 2 (0.37), RVE (25.69), and MAE (8.17) but the performance was improved by the application of BCM. The ECDF combined with the DM shows the best performance in stream flow simulation with NSE (0.63), R 2 (0.64), RVE (-0.43), and MAE (4.78).