Assessment of Hybrid Downscaling Techniques for Precipitation Over the Po River Basin

Although regional climate models (RCMs) are powerful tools for describing small scale climate conditions, their direct use in impact studies is still challenging since they are commonly biased. In order to provide "corrected'' climate scenarios, here we propose and compare three different hybrid downscaling methods for precipitation, that merge dynamical and statistical approaches, using in cascade the COSMO-CLM RCM and three post-processing techniques: linear-scaling, quantile mapping and MOS analogs. Comparison results over the Po river basin indicate that these approaches generally improve the direct outputs of the COSMO-CLM model, and, among the methods, better results have been obtained with the quantile mapping technique. Finally, the application of these methods to future scenarios is also discussed.