全球陆地暴雨时空格局变化的自然和人文影响因素研究(1981-2010年)

采用多种统计方法厘清了自然和人文因素对1981-2010年全球陆地暴雨变化的贡献大小.在时间上,首先从33个自然因子和11个人文因子中通过逐步回归筛选出对全球陆地年际暴雨雨量和雨日有影响的ENSO、AO2个自然因子和AOD、CO22个人文因子.其次采用平稳性检验、协整性检验和Granger因果检验,结果表明筛选的4个因子能够有效解释1981-2010年全球陆地年际暴雨雨量和雨日的动态变化.最后基于多元线性回归的方差解释率发现自然和人文因子对全球陆地年际暴雨雨量和雨日的总方差解释率达52.2%和51.3%;自然因子对全球陆地年际暴雨雨量和雨日的方差解释率为32.2%和31.4%,占总方差解释率的61.7%和61.2%;人文因子则仅占总方差解释率的38.3%和38.8%.水汽因素的诊断表明,全球水汽因子不是影响全球陆地暴雨动态变化的主导因素.在空间相关上,自然因子与全球陆地暴雨的空间相关性高于人文因子. 

Climatical and anthropogenic factors in changes in spatiotemporal patterns of global heavy rainfall (1981-2010)

Frequent occurrence of extreme precipitation events under the background of global climate change has caused tremendous harms to economic and social development, life security, and ecosystem,with profound impacts on sustainable development in disaster area. Heavy rain has become a key factor in global &. regional disasters,and has become a major concern in academic circles and in society. Stepwise regression analysis identified 2 out of 33 climate indices-El Nino-Southern Oscillation (ENSO),Arctic Oscillation (AO), and 2 out of 11 socio-economic factors-black carbon aerosol optical thickness (AOD) and carbon dioxide (C02), all closely related to urbanization, as important factors for increased heavy rainfall globally. Granger causality test shows that urbanization factors and climate indices are both strong causes for increased heavy rainfall in global land,climate indices play more important roles in heavy rainfall than anthropogenic factors. Urbanization factors could explain 20.0％, and 19.9％ of the variance in HRA and HRD respectively, while natural factors could explain 32.2％ and 31.4％ respectively. Distributions of HRA and HRD demonstrate spatially a clear expanding pattern, which show statistically significant correlation with spatial distribution of convective available potential energy. These data strongly suggest that significant increase in heavy rainfall across large area in global land during the past 30 years may have been mainly a consequence of large scale climate changes and rapid urbanization. All data suggest that large scale climate index variation is the main cause of annual and decadal increases in heavy rainfall. Our data also point out the importance of better understanding the role of human activities in regional climate change,and the importance of integration them into climate models.