气候变化条件下海河流域日降水统计降尺度研究(英文)  

Statistical Downscaling the Daily Precipitation for Climate Change Scenarios in Haihe River Basin of China

作　　者：褚健婷[1,2] 夏军[1] 许崇育[3] 

机构地区：[1]中国科学院地理科学与资源研究所陆地水循环与地表过程重点实验室,北京100101 [2]中国科学院研究生院,北京100049 [3]奥斯陆大学地球科学系,挪威

出　　处：《自然资源学报》2008年第6期1068-1077,共10页Journal of Natural Resources

基　　金：The Key Project of International Cooperation of the Natural Science Foundation of China(No.40730632);the Special Fund of Ministry fo Science & Technology,China(No.2006DFA21890);the Knowledge Innovation Key Project of the Chinese Academy of Sciences(No.KZCX2-YW-126)

摘　　要：在气候变化对水资源影响评价研究中,统计降尺度方法通常用来解决空间尺度不匹配问题。在论文中,预报量采用海河流域上典型区域的11个气象站点1961～2000年共40年的日降水资料,预报因子选自NCEP/NCAR的再分析资料,以及英国Hadlay中心GCM模型在a2情景和b2情景下的大气变量,主要采用Statistical Down-Scaling Model(SDSM模型)进行海河流域日降水的降尺度研究。结果表明:①使用逐步线性回归方法(SMLR),可以在广阔的空间范围内优选出具有一定物理意义的预报因子;②使用统计降尺度方法SDSM,采用不同的GCM资料,都可以较准确模拟降水的季节变化特征;③使用NCEP/NCAR再分析资料的降尺度结果优于Hadlay中心a2和b2情景下的,观测和模拟降水的相关系数分别可以达到93%和85%左右;④与现状年(1984～1993年)相比,未来(2011～2040年)的年降水总量大约会减少4%,降水峰值略滞后,降水时间更加集中,极端降水事件的强度加大,最大干旱持续天数也变长。Statistical downscaling methods are commonly used to deal with the spatial scale mismatch in assessment the impact of global climate change on water resources. In this article, the predictand is daily precipitation from 11 weather stations in Haihe River for 1961 -2000. The predictors originated from the reanalysis data set of NCEP/ NCAR, as well as the outputs of scenario a2 and b2 of HadCM3. The results showed that ： （ 1 ） predictors can be selected from the wide space with the Stepwise Multi-Line Regression method （ SMLR）, and these predictors have some physical meaning; （2） when the Statistical DownScaling Method （SDSM） was driven by different GCMs outputs, the annual precipitation cycles were well downscaled with all of the GCM reanalysis data ; （3） the performance of NCEP/NCAR data was overall better than the h3a2/h3b2, the correlation coefficients of area-average rainfall were 93% and 85% , respectively; （4） when compared with the current climate condition （1984 -1993 ） , in the future （2011 -2040）, the total amount of annual precipitation will decrease by about 4% ; the rain event may be more concentrated, and the intensity of extreme event can be reinforced, such as the maximum 1 - day, 5 - day rainfall, and the maximum length of dry spell will also be clearly lengthened.

关 键 词：气候变化 水文响应 统计降尺度方法 日降水 中国海河流域 SDSM方法 SMLR方法 

分 类 号：P467[天文地球—大气科学及气象学] P426.613