Future climatic and hydrologic changes estimated by bias-adjusted regional climate model outputs of the Cordex-Africa project: case of the Tafna basin (North-Western Africa)

International audienceThis study investigates climatic and hydrologic changes of the Tafna basin, by using ten outputs of precipitation and temperature from RCMs of the Cordex-Africa project. Different methods of bias-correction (LS, LOCI, DM and VS) are compared to correct the bias of precipitation and temperature datasets to observations. The suitable method, DM, reduces the bias to 1.27 mm for precipitation and 0.06 and 0.7°C for minimum/maximum temperature, respectively. The bias-corrected precipitation and temperature datasets are introduced into the SWAT model, calibrated and validated on the Tafna basin with good Nash criteria (NSEoutlet = 0.83). The discharge is over or under-estimated without bias-correction of RCM outputs, which highlights the necessity of applying bias-correction before using RCM outputs from Cordex-Africa for hydrological applications. The results show that the precipitation and discharge decreases, and temperature increases are more important with RCP 8.5 than with RCP 4.5, especially in the last decades of the 21st century