基于REOF-EEMD的西南地区气候变化区域分异特征

西南地区是全球变化区域响应的特殊地区,探究其气候变化区域分异特征具有重要的科学意义。文中选用REOF方法开展研究区气温和降水变化特征的空间分区,借助EEMD与BG分割算法等方法细致辨析了不同气候分区的气候演变特征。结果显示：① 西南地区年均温和年均降水变化均可划分为3个亚区,各自的空间界限高度相似,但降水Ⅱ、Ⅲ区的界限更偏南。② 20世纪50年代以来各气温亚区的年均温显著升高,川渝气温变化与全球变暖同步,黔西、黔中、滇北散布若干点状冷区。各降水亚区的时空差异明显,相较Ⅲ区,Ⅰ、Ⅱ区年均降水的波动性及年代际变化的差异更显著。③ ENSO事件对研究区气候变化的影响深远,不同气温、降水亚区对其的响应不尽相同。④ 不同气温亚区年均温序列突变点的收敛性较强,大致发生在1997年前后。不同降水亚区年均降水序列突变点的收敛性较弱。⑤ 各气温亚区年均温增加的持续性较强,Ⅱ、Ⅲ区尤甚。降水Ⅰ、Ⅱ区降水变化趋势不甚明显且具有一定的随机性,Ⅰ区的可能呈减速趋缓的减湿趋势,Ⅱ区的可能出现弱度减湿趋势,Ⅲ区降水趋于弱增。

Spatial division of climate change and its evolution characteristics in Southwest China based on REOF-EEMD

Southwest China is a special region in the context of global climate change. It is crucial to study the climate change in different parts of this region. In addition, it is important to explore the climate change characteristics of the counties in the study area which have benchmark meteorological stations. In this paper, the Rotated Empirical Orthogonal Function (REOF) was chosen to carry out climate regionalization according to annual mean temperature and annual mean precipitation change in Southwest China. Despite that, the Ensemble Empirical Mode Decomposition (EEMD), Bernaola Galvan heuristic Segmentation Algorithm (BGSA) and other methods were used to analyze the spatio-temporal evolution of climate change in subregions of the study area. Here are the results: (1) According to annual mean temperature and annual mean precipitation changes, the temperature and precipitation changes in the study area could be regionalized into three subregions respectively. The spatial boundaries of these two types of subregions were extremely similar, and the southern boundaries of precipitation subregions II and III were much more boarder than those of the temperature subregions II and III. (2) Annual mean temperatures in three temperature subregions have been rising significantly since 1950, and the temperature changes in Sichuan and Chongqing kept pace with global warming. It is worth mentioning that there were several cold areas sporadically distributed in the western Guizhou, the central Guizhou and the northern Yunnan. In terms of precipitation subregions, the spatial and temporal differences of annual mean precipitation were much larger, and the spatial and temporal differences of the interdecadal variation in the precipitation subregions I and II were more prominent than those in the precipitation region III. (3) The ENSO events had a profound influence on the climate change of Southwest China, and these three temperature and precipitation subregions responded differently to it. (4) The convergence of mutational site of annual mean temperature series in the three temperature subregions was strong (which all began around the year 1997), while that of annual mean precipitation series in the three precipitation subregions was relatively weak. (5) The climate warming trends in the three temperature subregions were obvious, especially in the subregions I and II. Regarding to annual mean precipitation, the trends of the precipitation in subregions I and II had certain randomness, the trend of annual mean precipitation in precipitation subregion I may decelerate with slower deceleration, and that in precipitation subregion II may weakly decelerate, while that of subregion III may weakly accelerate.