Projections In Extreme Climate Events In China During 1981–2100 Based On Global Climate Models

Climate change is the key point of quaternary environmental change.After the industrial revolution,the impact of human activities on global warming has deepened,accompanied by extreme climate events,which will have serious impacts on human society and the terrestrial ecosystem.The analysis and prediction of the spatial and temporal distributions of long-term extreme climate events will help climate risk assessment and disaster warning.Based on the observed data from 763 ground meteorological stations in China and the downscaled global climate model data from CMIP5 and CMIP6,the typical meteorological parameters and extreme climate events in China from 1981 to 2100 are analyzed in this study.We first analyzed the temporal and spatial distributions of air temperature,precipitation and surface solar radiation under different baseline periods and future scenarios,and then compared multiple potential evapotranspiration estimation models based on air temperature and surface solar radiation.At last,the temporal and spatial distributions and future changes of extreme cold/warm events,extreme precipitation events and drought events in China are predicted and analyzed,and the differences between CMIP5 and CMIP6 in the projections of future climatic parameters and extreme climate events are compared.The main conclusions are as follows:（1）During the baseline period,the statistical downscaled air temperature data provided by CMIP5 and CMIP6 had better consistency with the observed data than the precipitation data.During the baseline period,the maximum and minimum temperatures showed increase trends by 0.4℃/decade in China,especially in the northeast region;the precipitation mainly increased in northern China and decreased in the southeastern coastal areas,and the national average decrease trend is-7.6 mm/decade.The surface solar radiation mainly decreased in northern China and increased in northwestern China and the southeastern China,and the national average decline trend is-0.1MJ/m²/day/decade;（2）The Penman model based on the physical processes was used to estimate the potential evapotranspiration during study period.Then three simplified empirical models based on radiation and air temperature compared with the Penman model.The Abtew model will underestimate the potential evapotranspiration in China;the Hargreaves model will overestimate the potential evapotranspiration and the Jensen-Haise model will underestimate the potential evapotranspiration in the low-value area,but overestimate in the high-value area.The Nash coefficient（NSE）and root mean square error（RMSE）ranging from 0.76 to 0.88 and 16 to 23%for the three empirical models.（3）The uncertainties of the temperature and precipitation data under the future scenarios provided by CMIP5 models are higher than those of CMIP6,and the uncertainty of precipitation data is higher than that of air temperature data in same set of data sources.From 2011 to 2100,the median values of maximum and minimum temperatures provided by CMIP5 multi-model increase both by 0.2-0.5°C/decade,while in CMIP6 they were increase by 0.5-0.6 and 0.3-0.6°C/decade,respectively;The multi-model median increases in precipitation provided by CMIP5 and CMIP6 are10.4-14.3 mm/decade and 13.3-22.6 mm/decade,respectively.The changes in the surface solar radiation provided by CMIP5 and CMIP6 are relatively gentle in the future,and the changes in the multi-model median values??under future scenarios are lower than those during the baseline period.The potential evapotranspiration modeled by the Abtew and Jensen-Haise models increases more than the Hargreaves model under the future scenarios.For the long-term temperature,precipitation and potential evapotranspiration,the order of their changes amplitudes under different future scenarios is SSP585>RCP8.5>SSP245>RCP4.5.（4）15 ETCCDI climate change indices were used to study the spatiotemporal distributions of extreme climate events in China,including 3 annual temperature indices,6 extreme temperature indices and 6 extreme precipitation indices.The results showed that the extreme warm events in China are mainly higher in southern China and lower in northern regions during the baseline period,and the extreme cold events are on the contrary.Extremely high precipitation mainly occurs in southern China,while extreme drier event mainly occurs in northwestern China.During 1981–2100,extreme high temperature events showed upward trends in China,while extreme cold events showed downward trends.On the temporal and spatial scales,both the increase in extreme high temperature events and the decrease in extreme low temperature events will be the highest in northern China in the late 21st century,and the changes under the high emission scenarios in CMIP6 were larger than those under the other three scenarios.In addition,the extreme precipitation in China mainly increased in the southeastern region,but the 25–75th percentiles of the multi-model results of CMIP5 and CMIP6 were significantly different,indicating that different models have large differences in the prediction of future extreme precipitation events.（5）Based on the long-term precipitation data and the potential evapotranspiration data estimated by different models,the seasonal SPEI values in 763 stations in China were calculated.The drought events were divided into mild drought（-1<SPEI≤-0.5）,moderate drought（-1.5<SPEI≤-1）and severe drought（SPEI≤-1.5）.The results show that the spatiotemporal distributions of drought events based on different model data in baseline periods is relatively consistent:the mild droughts are dominant in spring,autumn and winter,and severe droughts are dominant in summer.From 2011 to 2100,the drought events in each season in China showed decrease trends during early 21^stcentury and then increase during late 21^stcentury compared with the baseline period.The variation of drought frequency under the high emission scenario was larger than that in the medium emission scenario,which was mainly due to the higher increase trends in potential evapotranspiration under the high emission scenarios than that under the middle emission scenarios.