台风诱发暴雨降水量的概率预测  

Probability Prediction of Typhoon-Induced Rainfall

作　　者：刘德辅[1] 谢波涛[1] 伍远康 庞亮[1] 

机构地区：[1]中国海洋大学工程学院防灾研究所,山东青岛266003 [2]浙江省水文局,浙江杭州310009

出　　处：《中国海洋大学学报（自然科学版）》2007年第6期1027-1033,共7页Periodical of Ocean University of China

基　　金：国家自然科学基金项目(50679076)资助

摘　　要：每年由台风诱发暴雨引起洪水、滑坡等灾害在我国沿海地区造成巨大损失。作者采用多维复合极值分布理论(Multivariate compound extreme value distribution),以受台风暴雨危害较严重的浙江温州地区为例,对台风诱发暴雨的降水量进行了概率预测。该理论以影响暴雨强度的台风中心气压差和降水量为研究对象,并考虑导致极端降水的台风过程的出现频次的概率特性,预测结果较传统方法更为合理和安全。Typhoon induced rainfall disasters, such as river floods and land slides, may cause serous damage in most coastal areas of China. This paper proposes a new probability model-Multivariate compound extreme value distribution（MCEVD） to predict typhoon-induced rainfall for the Wenzhou City in Zhejiang Province. In this new model, the typhoon central pressure drop, rainfall, and occurrence frequencies of extreme rainfall induced by typhoons are taken into consideration. It is shown that the new model can produce more reasonable predicted results than traditional methods.

关 键 词：台风 暴雨 降水量 多维复合极值分布 概率预测 

分 类 号：P426.6[天文地球—大气科学及气象学]