Simulation And Estimation Of Regional High Temperature Events Over China

For the problem of downscaling in daily maximum temperatures, a new statistical downscaling method named Optimal Points Regression (OPR) has been developed. It was uesd to downscale daily maximum temperature data from three models provided by IPCC AR5(CanESM、NorESM、BCC)and get muti-model ensemble results through the arithmetic average method. Downscaling effects and projected change of daily maximum temperature in china were discussed. On this basis, an Objective Identification Technique for Regional Extreme Events(OITREE) method was uesd to simulate and estimate regional high temperature events. Simulation effects and projected change of regional high temperature events in future were also analyzed. The main conclusions can be summarized as follows:(1) A new statistical downscaling method named Optimal Points Regression (OPR) has been developed. A comparison study on downscaling in daily maximum temperatures in the plain area (Shandong province) and the plateau area (Yunnan province) of OPR and another downscaling method (the Bilinear Interpolation, BI) was carried out. Results show that, both for plain area or plateau area and in summer (July) or winter (January), the OPR method is obviously better than the BI method.(2) Tested and compared the downscaling effects of three models and muti-model ensemble. Projected changes of daily maximum temperatures in the future over china were analyzed. The simulation effects of BCC model and muti-model ensemble are better than CanESM model and NorESM model. In the future scenario, maximum temperatures over china will increase, but have obvious regional difference.(3) Change characteristics of high temperature events in1961-2005and simulation effects were analyzed, results show that:there were157regional high temperature events in1961-2005, while frequence of the general events are the most, the weak events are the second,the severe events are the third, and the extreme events are the least. With the pass of time, regional high temperature events occur more frequently, more intense, and their influence range is wider. Regional high temperature events occur most frequently in July and August. The simulation effects of BCC model and muti-model ensemble are better then CanESM model and NorESM model.(4) Analyzed the projected change of regional high temperature events in future and made comparison between models and muti-model ensemble. Rising range of indices of CanESM model and BCC model is bigger then NorESM model and muti-model ensemble. Regional high temperature events will mainly occur in April to September over most regions of china in2006-2099.