| 广义极值分布在重庆短历时极值降水中的应用 |
| |
| 引用本文： | 王颖,刘晓冉,程炳岩,孙佳,廖代强.广义极值分布在重庆短历时极值降水中的应用[J].气象,2019,45(6):820-830. |
| |
| 作者姓名： | 王颖 刘晓冉 程炳岩 孙佳 廖代强 |
| |
| 作者单位： | 重庆市气候中心，重庆 401147,重庆市气候中心，重庆 401147,重庆市气候中心，重庆 401147,重庆市气候中心，重庆 401147,重庆市气候中心，重庆 401147 |
| |
| 基金项目： | 公益性行业(气象)科研专项(GYHY201506014、GYHY201306053)、重庆市气象局青年基金(QNJJ-201604)和重庆市建设科技计划项目［城科字2015第(2-10)号］共同资助 |
| |
| 摘 要： | 利用广义极值分布函数拟合1981—2016年重庆34个国家气象站短历时(1、3、6、12 h)极值降水序列,对拟合结果进行显著性水平检验,并给出不同重现期极值降水的空间分布。结果表明:广义极值分布函数能较好地拟合重庆地区的短历时极值降水。随着降水历时的延长,服从Weibull分布(Frechet分布)的站点数逐渐减少(增加)。各短历时不同重现期降水的空间分布具体表现为10 a以下及20 a以上基本相似,位于长江沿线以北的重庆西北部地区降水量明显大于重庆长江沿线以南地区,且渝东南降水的相对大值区位于彭水地区。随着重现期的增加,降水中心更加集中,渝东北的大值中心随着历时的延长向北移动。广义极值分布函数的形状参数的绝对值接近或超出0.5时,计算的高重现期(大于样本长度)极值降水存在较大偏差;当不同历时降水拟合的形状参数值具有明显差异时,高重现期降水可能出现与客观规律相悖的现象。
|
| 关 键 词： | 短历时降水，年最大值取样，广义极值分布，重现期降水 |
| 收稿时间： | 2018/6/10 0:00:00 |
| 修稿时间： | 2019/4/2 0:00:00 |
Application of Generalized Extreme Value Distribution Model to Short-Duration Extreme Precipitation in Chongqing |
| |
| WANG Ying,LIU Xiaoran,CHENG Bingyan,SUN Jia and LIAO Daiqiang.Application of Generalized Extreme Value Distribution Model to Short-Duration Extreme Precipitation in Chongqing[J].Meteorological Monthly,2019,45(6):820-830. |
| |
| Authors: | WANG Ying LIU Xiaoran CHENG Bingyan SUN Jia and LIAO Daiqiang |
| |
| Institution: | Chongqing Climate Center, Chongqing 401147,Chongqing Climate Center, Chongqing 401147,Chongqing Climate Center, Chongqing 401147,Chongqing Climate Center, Chongqing 401147 and Chongqing Climate Center, Chongqing 401147 |
| |
| Abstract: | Generalized extreme value (GEV) distribution is used to fit the short duration (1, 3, 6 and 12 h) extreme precipitation sequence of 34 national weather stations in Chongqing from 1981 to 2016. The results are as follows: GEV distribution can well fit the short-duration extreme precipitation in Chongqing. As the precipitation lasts longer, the number of the national weather stations subject to Weibull (Frechet) distribution gradually decreases (increases). The spatial distributions of different return periods are basically similar under 10 a and over 20 a. The precipitation to the north of Yangtze River, which is located in the northwestern part of Chongqing, is obviously more than in the areas to the south. And in southeastern Chongqing, there is more precipitation in Pengshui. The precipitation centers are more concentrated with the increase of recurrence period, and the great value center in Northeast Chongqing moves northward as time goes on. When the absolute value of the shape parameter of GEV distribution function is close to or greater than 0.5, the extreme precipitation of the calculated high recurrence (longer than the sample length) is greatly deviated. If there exists prominent difference in the shape parameter of different-duration precipitation, the result appears to be contrary to objective reality. |
| |
| Keywords: | short-duration precipitation year maximum method generalized extreme value (GEV) distribution return period rainfall |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《气象》浏览原始摘要信息 |
| 点击此处可从《气象》下载免费的PDF全文 |
|
