资源环境科技发展态势分析平台

Dynamical downscaling of climate projections over a limited-area domain using a Regional Climate Model (RCM) requires boundary conditions (BC) from a Coupled Global Climate Model (CGCM) simulation. Biases in CGCM-generated BC can have detrimental effects in RCM simulations, so attempts to improve the BC used to drive the RCM simulations are worth exploring. It is in this context that an empirical method involving the bias correction of the sea-surface conditions (SSCs; sea-surface temperature and sea-ice concentration) simulated by a CGCM has been developed: The 3-step dynamical downscaling approach. The SSCs from a CGCM simulation are empirically corrected and used as lower BC over the ocean for an atmosphere-only global climate model (AGCM) simulation, which in turn provides the atmospheric lateral BC to drive the RCM simulation. We analyse the impact of this strategy on the simulation of the African climate, with a special attention to the West African Monsoon (WAM) precipitation, using the fifth-generation Canadian Regional Climate Model (CRCM5) over the CORDEX-Africa domain. The Earth System Model of the Max-Planck-Institut fur Meteorologie (MPI-ESM-LR) is used as CGCM and a global version of CRCM5 is used as AGCM. The results indicate that the historical climate is much improved, approaching the skill of reanalysis-driven hindcast simulations. The most remarkable effect of this approach is the positive impact on the simulation of all aspects of the WAM precipitation, mainly due to the correction of SSCs. In fact, our results show that proper sea surface temperature (SST) in the Gulf of Guinea is a necessary condition for an adequate simulation of WAM precipitation, especially over the equatorial region of West Africa. It was found that the climate-change projections under RCP4.5 scenario obtained with the 3-step approach are substantially different from those obtained with usual downscaling approach in which the RCM is directly driven by the CGCM output; in the WAM region most of the differences in the projected climate changes came mainly from the empirical correction of SST.