面向灾害风险评估的热带气旋路径及强度随机模拟综述

历史热带气旋记录时间序列较短空间差异大,热带气旋灾害风险评估经常面临样本不足,特别是超强台风及巨灾记录历史样本的问题,从而导致传统概率统计方法失效。过去20多年来,逐渐发展出一套完整的方法体系进行热带气旋路径及强度随机模拟,其特点是充分利用历史总体样本信息,生成大量符合历史样本特征的热带气旋路径及强度随机事件样本集,从而有效地解决了局地历史样本不足的问题。在回顾热带气旋的年频次、季节分布、路径分布、强度及影响范围时空规律研究进展基础上,系统综述了用于热带气旋路径及随机模拟的起始点模型、行进模型、终止点模型、洋面强度模型、陆地衰减模型及结果检验方法等领域的进展及不足,然后对其在世界各地的应用进行了概述,并对未来研究改进方向及应用领域进行了展望。

A Review of Stochastic Modeling of Tropical Cyclone Track and Intensity for Disaster Risk Assessment

The samples of Tropical Cyclone (TC) track and intensity play an essential role in TC disaster risk modeling. However, inadequate historical records may introduce great bias into risk assessment results if traditional probabilistic method is used. In order to solve the challenge, the methodology for developing stochastic TC events have been proposed and improved greatly in the past 20 years based on the understanding of the spatiotemporal characteristics of TC observations. By generating a large amount of stochastic TC events, the methodology is able to make full use of historical TC information. In this paper, the latest progress of the analysis of TC spatiotemporal characteristics, and the methods for the stochastic modeling of annual occurrence, seasonal distribution, genesis, track forward velocity and direction, decay and intensity of TC are systematically reviewed. The applications of the methodology for tropical risk assessment in different ocean basins and countries are also outlined. In addition, the potential of the methodology in the fields of risk assessment, insurance, design code, and emergency planning are also discussed.