Comparison between past and future extreme precipitations simulated by global and regional climate models over the Tibetan Plateau

Past studies on regional climate change over the Tibetan Plateau (TP) have mainly looked at changes in the mean climate. This study focuses on past and future extreme precipitations, simulated by global and regional climate models over the TP. To assess the influence of large‐scale forcing on dynamic downscaling using Weather Research and Forecasting (WRF) model, downscaling results for the historical period (1980–2005) with ERA‐Interim reanalysis and CCSM4 as forcings are evaluated against a gridded observational data set. These are inter‐compared before future projections for the period 2005–2100 under two scenarios (RCP4.5 and RCP8.5). The followings are obtained: (1) the reanalysis greatly overestimates not only the mean, but also extreme precipitation. The overestimation in CCSM4 is even larger than that of the reanalysis. (2) The two downscalings outperform their forcings, reflected by reduced overestimation for extreme precipitation frequency, increased spatial pattern correlations and more accurate linear trends, especially for the downscaling driven by CCSM. This demonstrates the constraining power of the fine‐scale modelling and the importance of more realistic representations of surface forcing and related processes in the TP. (3) CCSM4 projects a general wetting across the whole TP with increases of heavy precipitation as well as the wetting intensification with warming. WRF also projects an overall wetting, but the wetting is less sensitive to the warming and there is more of an increase in light precipitation frequency. More importantly, a diverse pattern with wetting in the north and drying in the south is found in the dynamical downscaling in contrast to the uniform wetting in its forcing.