Bayesian multi-model projections of extreme hydroclimatic events under RCPs scenarios

Abstract

A Bayesian multi-model inference framework was used to assess the changes in the occurrence of extreme hydroclimatic events in four major river basins in China (i.e., Liaohe River Basin, Yellow River Basin, Yangtze River Basin, and Pearl River Basin) under RCP2.6, RCP4.5, and RCP8.5 scenarios using multiple global climate model projections from the IPCC Fifth Assessment Report. The results projected more summer days and fewer frost days in 2006–2099. The ensemble prediction shows the Pearl River Basin is projected to experience more summer days than other basins with the increasing trend of 16.3, 38.0, and 73.0 d per 100 years for RCP2.6, RCP4.5 and RCP8.5, respectively. Liaohe River Basin and Yellow River Basin are forecasted to become wetter and warmer with the co-occurrence of increases in summer days and wet days. Very heavy precipitation days (R20, daily precipitation ≥20 mm) are projected to increase in all basins. The R20 in the Yangtze River Basin are projected to have the highest change rate in 2006–2099 of 1.8, 2.5, and 3.8 d per 100 years for RCP2.6, RCP4.5 and RCP8.5, respectively.