Projection and uncertainty of precipitation extremes in the CMIP5 multimodel ensembles over nine major basins in China

摘要

This study presented an analysis of projection and uncertainty of precipitation extremes over 9 river basins in China based on the outputs of 18 global climate models (GCMs) with 3 Representative Concentration Pathways (RCPs) from the Coupled Model Inter-comparison Project Phase 5 (CMIP5). The temporal and spatial changes in precipitation extremes for the period 2020-2100 were analyzed using the climate indices defined by the Expert Team on Climate Detection and Indices (ETCCDI). Uncertainty of GCMs and RCPs in the projections of precipitation extremes was quantified using the variance-based sensitivity analysis method. The model simulations generally predict a consistent intensification of precipitation extremes during the 21st century with respect to different RCPs scenarios: the higher emission scenarios (e.g., RCP8.5), the larger intensity and frequency of precipitation extremes. The projected changes in precipitation extremes exhibit a large spatial variation across China. The high-latitude and high elevation regions of China (e.g., Continental and Southwest basins) are projected to respond more strongly to the increase in precipitation amount and intensity (e.g., RCPTOT, RX1day and RX5day), while southeastern and southern China (e.g., Southeast and Pearl River basins) tend to be more sensitive to the increase in the frequency of precipitation extreme (e.g., R10mm and R20mm). Consecutive dry days (CDD) is projected to decrease in northern China (e.g., Continental basin) but increase in southern China (e.g., Southeast, Pearl River, and Yangtze River basins). Uncertainty analysis shows that variation of GCMs contributes 90% uncertainty of precipitation extremes projections for the whole China, with larger uncertainty range under the higher emission scenarios (e.g., RCP8.5). The uncertainty from RCPs is generally limited (contribution 10%) in comparison with GCMs, and slightly increases with the long-term projections of precipitation extremes.