Estimation of Extreme Probability Distribution Tails: Application to Extreme Wind Speeds

该项目将使用极值统计技术来研究风速数据的上尾部分布。极端事件概率分布的尾部估计是一个重要的理论和工程问题。飓风、龙卷风和其他强风事件每年造成生命损失和实质性的经济和结构破坏，如果加强对极端情况的预测，这些损失就可以减少。这项工作的目的是提高极端风预测能力，缩小观测到的损害统计数据与使用传统分布校准程序获得的损失统计数据之间的差距。目前的设计风速是基于对大约100个气象站几十年的天气服务数据的分析。一个更新的数据库将被用来检验分布具有有限上尾的假设。将对所作假设的物理基础进行检验，然后对条件超越进行校准，并对这一统计方法进行验证。

The project will use extreme value statistical techniques to study the upper tail distribution of wind speed data. The estimation of the tails of probability distributions of extreme events is a theoretical and engineering problem of significance. Hurricanes, tornadoes and other high winds episodes produce every year loss of life and substantive economic and structural damages which could be curtailed if prediction of extremes would be enhanced. The purpose of this work is to improve the extreme wind prediction capabilities and to close the gap between the observed damage statistics and those obtained by using conventional distribution calibration procedures. Current design winds speeds are based on the analysis of few decades of weather service data at about 100 weather stations. An updated data base will be used to test the hypothesis that the distribution has a limited upper tail. The physical basis for the assumptions made will be tested, followed by a calibration of conditional exceedance and validation of this statistical approach.