中国近1000年旱涝的持续性特征研究

应用中国近531年旱涝指数和近1041年干湿指数，定义干旱(湿润)等级，滑动计算原指数序列各干旱(湿润)等级的出现次数，发现各等级出现的次数与其窗口长度之间均遵循指数分布P_i(x)=Ae^-γx.结合指数分布的数理意义，定义指数特征值γ的倒数λ为描述旱(涝)持续性的尺度因子并对华北和江淮流域旱(涝)发生的持续性特征进行研究.结果表明：旱尺度因子的空间分布表现为由北向南呈带状式波动分布，我国北方地区干旱的持续性相对长江流域要长一些，由北向南三个区的旱尺度因子的均值分别为187，162，182.旱涝指数序列中旱或偏旱(涝或偏涝)相对集中的时段对旱(涝)持续性影响较显著，12世纪末期、13世纪早期、17世纪早期和20世纪末期华北和江淮流域发生时间上同步、空间上尺度较大的极端干旱事件的概率较高，这也从侧面验证了旱涝指数序列中群发现象的存在；华北地区1260—1280年(对应的气候背景为中世纪暖期的末期)的旱涝指数对旱尺度因子的影响较1980—2000年(对应的气候背景为20世纪全球增暖)的情况更显著；1260—1280年这一时段旱或偏旱年数较1980—2000年也要多一些.因此，在气候较暖的时期可能易发生强度大、范围广的同步干旱事件，而近30年的中国北方干旱化可能是自然变率起主导作用下人为变率和自然变率共同作用的结果. 关键词： 旱(涝)尺度因子 持续性 群发性 指数分布

The research of durative characteristics of dry/wet series of China during the past 1000 years

Using dry/wet index series of China during the past 531 years and 1041 years, by defining the dry (wet) grade and making running calculation of the appearance times of each dry (wet) grade in the original index series, it was shown that there is a exponential distribution P(x)=A^*e^-γx of the appearance times of each grade with respect to the window size. We defined 1/γ as the scale factor to characterize the dry (wet) durative and used it to analyze the dry (wet) durative of north China and the area between Yangtze and Huai rivers. Results show that the dry factor has the semi-belt wave distribution from north to south, the dry durative in north China is somewhat longer than that between Yangtze and Huai rivers, the average dry factors of area A, B and C are 1.87, 1.62 and 1.82, respectively. The dry (wet) durative is rather obvious in those intervals in which dry or wet are comparatively concentrated, so the large spatial scale and synchronized extreme dry accidents are easier to happen during the end of 12th century, the beginning of 13th century, the beginning of 17th century and the end of 20th century, which verifies the existence of cluster phenomena in the dry/wet index. The influence to dry durative in north China caused by the dry/wet index during 1260—1280 (the climate back ground being the end of the middle ages warm) was much lavger than that caused by the index during 1980—2000 (the climate background being the global warming). Furthermore, the number of dry years during 1260—1280 was also larger than that of 1980—2000. Obviously, serious, larger spatial scale and synchronized extreme dry accidents are easy to happen in warm periods, and the drought since 1970s in north China might be caused by co-operation of human and natural factors, in which the natural factor might have played the more important role.