基于多源数据的台风风暴潮概率预报研究:台风集合的构建

建立了一套用于台风风暴潮集合预报的台风集合构建方案。首先基于中国中央气象台、中国香港天文台、中国台湾中央气象局、美国联合台风预警中心、日本气象厅和韩国气象台6家预报中心的预报数据,构建一个误差更小的24 h、48 h和72 h预报时效的台风分析数据;然后基于分析数据构建9个路径样本(1条分析路径+2个概率圆上的8条概率路径)和3个台风最大风速(概率偏弱、居中和偏强)样本,形成27个台风样本集合,并根据分析风场的误差分布合理确定不同台风样本的发生概率。通过对台风"利奇马"的应用,证实该集合方案可以覆盖大部分可能的情景,集合样本具有较强的代表性,可用于台风风暴潮的集合预报。

Probabilistic forecast for typhoon storm surge based on multi-source data:Creation of typhoon ensemble

This study establishes a scheme for creating a typhoon ensembles that is used in storm surge probabilistic forecast.Based on the forecast products of the China Meteorological Administration,Hong Kong Observatory,Central Weather Bureau,Joint Typhoon Warning Center,Japan Meteorological Administration and Korea Meteorological Administration,a typhoon analysis dataset with higher accuracy for the 24-hours,48-hours and 72-hours forecast is created.The dataset is used to generate 9 members for track(1 analysis track and 8 probabilistic track on the probability circle)and 3 members for typhoon maximum wind speed(weak,medium and strong probability),which results in 27 typhoon ensemble members.Meanwhile,the probability of each ensemble members is determined based on the error distribution of the typhoon analysis dataset.The scheme is validated by typhoon"Lekima".The results show that the scheme includes most possible scenarios and the representativeness of the ensemble members are significant,which approves the applicability of the scheme in typhoon storm surge probabilistic forecast.