Simulation of future scenarios of climate change in the middle and upper reaches of the Peal River using the Statistical Down Scaling Model (SDSM)

XU Yan; WANG Shijie; BAI Xiaoyong; LI Xiongyao; SHI Xiaoming; TIAN Yichao; WU Luhua

doi:10.11932/karst20180209

Abstract：Due to the vulnerability and sensitivity of the karst eco-system,the prediction of future climate change in the karst area has great significance to reasonably exploit the regional resources and effectively pro-tect the ecological environment.However,the research on the climate downscaling in the karst area is nearly blank.In this paper,the daily mean temperature and daily precipitation from 13 meteorological stations in the Hongliujiang zone of the Pearl River Basin during 1961-2001 were used as the predictors.Moreover,the NCEP reanalysis data was used as the forecast factor.The SDSM model was applied to predict the trends of temperature and precipitation in future years under SRES A2 and B2 emission scenarios of the HadCM3 mod-el.The results show that,(1)SDSM mode can be applied in the accurate simulation of the temperature and precipitation in the study area,and the deterministic coefficient of observation and simulation value is about 99％ and 65％ respectively;(2)Compared with the reference period,the future temperature and precipitati-on in the A2 and B2 scenarios show a clear upward trend until the end of the 21 th century,and the increment is increasing with time prolonged.Meanwhile,by the end of 21th century,the relative variation of the future mean annual temperature in the A2 and B2 scenarios are + 3.39 ℃ and + 2.49 ℃ respectively,and the 117.30 ％ and 80.90 ％ growth respectively in daily precipitation under the A2 and B2 scenarios;(3)The rise of temperature is the most obvious in autumn and spring,and the precipitation increase is the largest in summer.Therefore,this analysis not only provides the foundation data and the theoretical basis for setting strategies,but also has great significance for impact study of climate changes in karst areas.