地理学报 ›› 2011, Vol. 66 ›› Issue (9): 1200-1210.doi: 10.11821/xb201109005

• 气候变化 • 上一篇    下一篇

近50 年来淮河流域极端降水的时空变化及统计特征

佘敦先1,2, 夏军1, 张永勇1, 杜鸿3   

  1. 1. 中国科学院地理科学与资源研究所陆地水循环及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. 武汉大学水资源与水电工程科学国家重点实验室, 武汉 430072
  • 收稿日期:2011-05-16 修回日期:2011-06-07 出版日期:2011-09-20 发布日期:2011-11-04
  • 通讯作者: 夏军(1954-), 男, 研究员, 博士生导师, 中国地理学会会员(S110001624M)。E-mail: xiaj@igsnrr.ac.cn E-mail:xiaj@igsnrr.ac.cn
  • 作者简介:佘敦先, 男, 安徽芜湖人, 博士生, 主要从事气候变化和极端事件研究。E-mail: shedunxian@sina.com
  • 基金资助:

    国家重点基础研究973 项目(2010CB428406); 国家自然科学基金项目(41071025)

The Trend Analysis and Statistical Distribution of Extreme Rainfall Events in the Huaihe River Basin in the Past 50 Years

SHE Dunxian1,2, XIA Jun1, ZHANG Yongyong1, DU Hong3   

  1. 1. Key Laboratory of Water Cycle and Related land Surface Processes, CAS, Beijing 100101, China;
    2. Graduate University of Chinese Academy of Sciences, Beijing 100049, China;
    3. State key laboratory of Water Resources and Hydropower Engineering Science, Wuhan University, Wuhan 430072, China
  • Received:2011-05-16 Revised:2011-06-07 Online:2011-09-20 Published:2011-11-04
  • Supported by:

    National Basic Research Program of China, No.2010CB428406; National Natural Science Foundation of China, No.41071025

摘要: 以淮河流域27 个气象站点1960-2009 年逐日降水观测资料为基础,选取年最大降水量序列(AM) 和超门限峰值序列(POT),分析淮河流域年极端降水事件的时空变化趋势,研究淮河流域降水极值的统计特征。研究发现:过去50 年,淮河流域大多数站点年最大日降水量有增加的趋势,少数站点有减少的趋势,但增加和减少的趋势均不明显。从单个气象站点50 年降水序列来看,年最大日降水事件发生的时间大多集中于20 世纪60-70 年代,且以汛期居多。利用L-矩法、K-S 检验等方法,发现GEV和GP分布分别能够较好的拟合AM和POT序列。通过计算比较在不同重现期水平下的降水量,发现POT序列及其对应的GP分布能够更好的模拟淮河流域极端降水序列。

关键词: 淮河, 极端降水, L-矩法, K-S检验

Abstract: Based on the daily precipitation data from 27 meteorological stations in the Huaihe river basin during 1960-2009, we analyzed the trend and statistical distribution of extreme precipitation events in this area. The annual maximum series (AM) and peak over threshold series (POT) are selected to describe the extreme distribution. The results show that positive trend of annual maximum precipitation is detected in most of the stations, only a small number of stations are found to be negative during the past five decades, and neither of the positive or negative trend is significant. The daily maximum precipitation events during a year almost occurred in the flood seasons of the 1960s-1970s. By the L-moments method, the parameters of three extreme distributions, which are GEV, GP and Gamma, are estimated. From the results of goodness of fit test and K-S test, we found that AM series can be better fitted by GEV model and POT series can be better fitted by GP model. In comparison with the precipitation amounts at different return levels, we found that the values obtained from POT series is slightly larger than the values fromAM series, and they can better fit the observed values in the Huaihe River Basin.

Key words: Huaihe River, extreme precipitation, L-moments method, K-S test