Future Prediction And Reliability Evaluation Of Regional Extreme Climate In China Based On Model Performance And Independence

Based on the simulations of the global climate models(GCMs)from the Coupled Model Intercomparison Project Phase 6(CMIP6),evaluated the simulation performance of the Climate Model Weighting by Independence and Performance(Clim WIP)ensemble scheme on China’s regional climate indices;the reliability of future estimates of ensemble schemes is discussed comprehensively using the “Perfect model test” framework,and the impact of different constraint indicators on regional climate estimation was explored;Finally,an integrated prediction of the future climate in China was conducted based on the selected constraint indicators.The main conclusions are as follows:(1): The ClimWIP scheme significantly reduced the cold bias of the temperature indices on the Tibetan Plateau and the wet bias of the precipitation indices on the periphery of the Tibetan Plateau in the historical period.The spatial correlation coefficient between the temperature indices and the observation field is above 0.98,and the relative standard deviation is close to 1.The spatial correlation coefficient of total precipitation and heavy precipitation with the observation field exceeds 0.92,and the relative standard deviation is between 0.8 and 1.0.(2): Based on the ClimWIP scheme,the ensemble prediction of future climate change in China at 1.5/2 ℃ global warming under SSP5-8.5 scenarios was carried out.The results showed that the projection uncertainty by the Clim WIP scheme is reduced compared with the unweighted scheme,the uncertainty of temperature of Southern China and Tibetan Plateau was reduced,and the uncertainty of precipitation in Northeast China and Northwest Xinjiang was significantly reduced,which could be reduced by more than 80%.Compared with the current period(1995-2014),the temperature indices increase predicted by the Clim WIP scheme is mainly concentrated in northern China and Tibetan Plateau,and the precipitation indices increases most significantly in the Northwest and Tibetan Plateau.Under an additional0.5 ℃ global warming,China’s regional average temperature indices would increase by more than 0.7 ℃,with greater warming in the Northeast and Northwest;total precipitation and heavy precipitation will also further increase.Under the global warming of 2 ℃,the warming magnitude in most regions of China would be likely larger than 1.5 ℃ compared to the current climate state(probability value > 50%).For the precipitation indices,the probability of wetter condition in Northwest China and North China would be larger,with a likely response magnitude exceeding 10% and 25% for PRCPTOT and R95P(probability value >50%).(3): Using the large ensemble mode with different climate sensitivities initial conditions as pseudo-observation,the “perfect model test” framework was constructed,and the reliability of the Clim WIP scheme based on different constraint indicators was evaluated for the future prediction of China’s regional extreme climate indices.The results show that considering more constraint indicators in the ensemble scheme can improve the reliability of future estimation results of the ensemble scheme.For maximum temperature(TXx),the Clim WIP scheme based on composite indicators that comprehensively consider historical climatic states and regional long-term trends is more reliable.Compared with the MME scheme,the Clim WIP scheme reduces the median deviation of the pseudo-observation simulation by more than 0.15 ℃,and the 5-95% uncertainty range is reduced by up to 1.25 ℃.Spatially,the effect is better south of the Yangtze River,and the uncertainty range is reduced by more than1.0 ℃.For the maximum consecutive 5-day precipitation(RX5day)in Eastern and Western China,the future projections of the Clim WIP scheme based on the composite index of the historical climate state of RX5 day and the long-term trend of the mean temperature region are more reliable.In Eastern China,compared with the MME scheme,the median deviation of the Clim WIP scheme for the pseudo-observation simulation can be reduced by up to 3.79%,and the uncertainty range can be reduced by more than 7.99%.The estimated regional average maximum temperature in China based on composite indicators increased by 2.18 ℃ and5.07 ℃ in the middle and end of this century compared with the current period(1995-2014),respectively,and the uncertainty decreased by 58% and 36%,respectively,compared with the MME scheme.The average RX5 day in the East and West regions of China increased by20.5% and 33%,respectively,at the end of the century compared to the current period,and the uncertainty was reduced by 23% and 18%,respectively,compared to the MME scheme.