热带气旋(Tropical Cyclone,TC)是一种生成于热带洋面上具有灾害性的天气系统,易诱发强降水、风暴潮等自然灾害现象,严重威胁沿岸地区人口生命和财产安全。因此,开展热带气旋危险性分析对于政府部门防灾减灾以及保险行业管理决策具有重要意义。由于热带气旋历史数据量不足,基于历史气旋样本直接开展气旋的危险性估计具有一定的不确定性。因此,在对气旋进行统计分析的基础上,利用随机模拟方法生成足量的气旋代用样本,不仅有助于揭示气旋活动的基本规律,还可以用于定量精细化的风险评估。本文研究目标为通过分析西北太平洋热带气旋时空分布规律,建立符合西北太平洋分布特征的热带气旋随机模拟,为风险管理及风险分散提供模型支持。本研究以IBTrACS(International Best Track Archive for Climate Stewardship)机构提供的最佳路径数据集作为模型输入气旋数据,随机模拟热带气从生成到消亡过程中路径及强度变化。具体研究内容和结论如下:
(1)分析西北太平洋热带气旋关键参数分布特征,利用经验性方法构建西北太平洋热带气旋统计模拟模型,随机模拟热带气旋从生成到消亡过程中路径及强度的变化,完成热带气旋随机事件集的构建,结果表明模拟气旋样本的检验结果与历史气旋样本基本吻合,可进一步应用于研究区热带气旋的定量精细化的风险评估中。
(2)分析大尺度环境因子与西北太平洋热带气旋生成活动的统计关系,完成基于物理参数的热带气旋生成模型的构建,随机模拟热带气旋起始点。结果表明模拟气旋生成点与历史气旋生成点分布区域基本一致,模型模拟结果稳健可靠。
以上研究结果表明,本研究构建的全路径重采样模型和生成回归模型具备描述热带气旋活动的能力,能为热带气旋危险性评估提供技术支撑。
本研究的创新点包括以下两点:(1)改进现有的非参数化模型,根据西北太平洋热带气旋的关键参数特征建立合适的经验全路径模型,提高运算效率。(2)考虑历史样本分布随机性,纳入热带气旋生成频次的趋势性和周期性,细化生成模型假设,优化生成模型,发展实际应用的可能性。
其他摘要
A tropical cyclone (TC) is a disastrous weather system that originates over warm tropical oceans, accompanying devastating phenomenon known as heavy rain and the storm surge, which pose the greatest threat to life and property. Therefore, it is of great significance to carry out risk analysis for disaster prevention & mitigation in government departments and decision making in the insurance industry. However, historical tropical cyclone are usually insufficient in quantitative TC risk analysis, making assessment of the risk difficult. One way to overcome this limitation is to make use of entire historical TCs, from genesis to lysis, thereby enhancing by roughly two orders of magnitude the amount of data on which to construct a statistical model. That’s helpful to reveal statistical law for tropical cyclone activity and develop quantitative risk assessment. The goal of this study is to develop a stochastic simulation model to generate enough TC samples in conformity to the spatiotemporal distribution of historical TC events in Northwestern Pacific. This study uses the best track dataset provided by IBTrACS (International Best Track Archive for Climate Stewardship) as the model input data, and randomly simulates the track and intensity changes of TC from genesis to lysis. Research contents and main conclusions are listed as follows:
(1) based on the historical observation TC data, spatiotemporal characteristics were analyzed and a basin-wide TC model was developed in this paper in order to construct datasets consisting of synthetic tropical cyclone tracks. By comparing simulations to the historical record using several diagnostics, the simulations and observations are statistically indistinguishable, which can be applied to quantitative TC risk assessment.
(2) statistical relationship between large-scale environmental factors and TC genesis activity in the Northwestern Pacific was analyzed and a genesis model was developed and optimized in this paper to randomly simulate TC genesis point. By comparing the spatial distribution of simulated sample and historical record, the simulations and observations are basically consistent, which show that the model is robust and reliable.
The above research results show that the full-track resampling model and genesis regression model constructed in this paper can describe tropical cyclone activities. It can provide technical reference for TC risk assessment and management.
The innovations of this study include the following two points: (1) Improve the existing non-parametric model and develop a basin-wide TC model (2) Optimize the genesis model.
修改评论