Methodological Approach for Climate Simulations Selection for Climate Change Impact Studies

Abstract This study builds on previous studies and proposes a combined approach for shortlisting and selecting representative climate simulations for impact studies in the Souss watershed in Morocco. The selection procedure consists of selecting climate simulations on the basis of predicted changes in means and extremes, climate model performance in simulating past climate and bias correction. Unlike the selection methods proposed in the literature that begin with a pre-selection of simulations based on changes in means, the approach recommended in this study involves an initial pre-selection of simulations by the ability of the models to simulate the base period climate. The proposed procedure involved selecting a set of general circulation models (GCMs) from the Phase 5 Coupled Model Intercomparison Project (CMIP5) and regional climate models (RCMs) from the Coordinated Regional Downscaling Experiment (CORDEX) project to project spatiotemporal changes in precipitation and temperature for the RCP4.5 and RCP8.5 scenarios. Climate changes between 1975–2005 and 2066–2095 were analyzed for the watershed. By 2066–2095, precipitation is projected to decrease by -12.2 to -30.1% under the RCP4.5 scenario and by -5.2 to -65.4% under the RCP8.5 scenario. Mean temperature is projected to increase by 2.3 to 2.6°C under RCP4.5 and by 4.4 to 4.7% under RCP8.5. For both scenarios, a decline in precipitation is projected while temperature is expected to increase. The Palmer Drought Severity Index (PDSI) revealed that the Souss watershed, characterized by incipient drought during the historical period, will experience mild and moderate drought in 2066–2095 under the RCP4.5 and RCP8.5 scenarios, respectively.