Comparative study of seven bias correction methods applied to three Regional Climate Models in Mekrou catchment (Benin, West Africa)

Abstract

An international awareness against climate change and its consequences is observed in recent decades. To predict adaptation to these climate changes, simulations of past and future climate were made using the Regional Climate Models (RCMs). These simulations are subject to bias; and many methods are developed to reduce these bias. In this study, seven (07) different methods (Delta change, Scaling, EQM, AQM, GQM, GPQM and ISIMIP) were applied to correct the precipitation of three (03) RCMs (REMO, DMI-HIRHAM5 and RCA4). Three (03) correction methods (Scaling, EQM and AQM) gave the most satisfactory results at different time scales (daily, monthly and yearly). The analysis of the future evolution of annual rainfall amounts for REMO model showed a downward trend for RCP4.5 and RCP8.5 scenarios with an emphasis on RCP8.5 scenario. As against the HIRHAM5 and RCA4 models, there was a tendency to increase on the evolution of annual rainfall amounts. The combination of the three models revealed a rising trend of future annual rainfall amounts for RCP8.5 scenario while the trend was almost constant for RCP4.5 scenario.