EAGER: Data-driven Physical Model for Hurricanes' Intensity-size Relation

EAGER：飓风强度-大小关系的数据驱动物理模型

• 批准号: 2012479
• 负责人: Guosheng Liu
• 金额: $ 14.99万
• 依托单位: Florida State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-03-15 至 2022-02-28
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2012479&HistoricalAwards=false
• 关键词: EAGER; Data; driven; Physical; Model

Hurricanes are among the most deadly and destructive storms impacting the United States, causing losses of life, devastating damages to buildings and infrastructures, and enormous financial losses. Both hurricanes’ intensity and size are the key factors determining their severity and destructive capability. Therefore, accurate prediction of hurricanes’ intensity and size is essential to general public and government officials. Observed hurricanes exhibit very rich and complex intensity-size relations. Even after taking the differences in the radius of the maximum wind into consideration, hurricanes with the same maximum wind can still have various sizes or hurricanes with the same size can have a large range of intensity. The existing empirical and theoretical models, however, tend to predict a nearly one-to-one relation between hurricanes’ intensity and size after taking the differences in the radius of the maximum wind into consideration. The goal of this project is to develop a data-driven physical model whose solutions can reproduce the rich and complex intensity-size relations of observed hurricanes. A successful project will lead to a better understanding of the physics governing hurricanes’ intensity-size relations. This project will provide a powerful tool to identify the major deficiencies in reproducing the rich and complex intensity-size relation by operational forecast models, leading to an improvement in forecasts and public safety and economic benefits. This project will train a postdoc in fields of atmospheric dynamics and data science and support STEM education by working with two undergraduate students on their honor theses research. The PIs will actively engage with North Florida Chapter of the American Meteorological Society and provide our experimental real time assessment of hurricanes’ intensity to the Florida State University/Tallahassee communities. This project will take a novel approach that combines theories and data-driven techniques to build a new model for the hurricanes’ intensity-size relation. Specifically, the PIs will utilize data analysis techniques to explore what are the factors controlling the variation of inward radial velocity among different hurricanes and then link these factors to the variation of hurricanes’ angular momentum loss. Another novelty of this project is that the developed model will be validated by examining its ability to predict the radial profile of azimuthal wind both inwardly from the boundary condition at outer radii and outwardly from the boundary condition at inner radii. Such exchangeability of the predictends and predictors allows ones to test whether the data-driven model possesses the quality of physical laws.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

飓风是影响美国的最致命和最具破坏性的风暴之一，造成生命损失，对建筑物和基础设施造成毁灭性破坏，并造成巨大的经济损失。飓风的强度和大小是决定其严重程度和破坏能力的关键因素。因此，准确预测飓风的强度和大小对公众和政府官员来说至关重要。观测到的飓风表现出非常丰富和复杂的强度-大小关系。即使考虑到最大风半径的差异，相同最大风的飓风仍然可以有不同的规模，或者相同规模的飓风可以有很大的强度范围。然而，现有的经验和理论模型在考虑到最大风半径的差异后，倾向于预测飓风的强度和大小之间几乎是一对一的关系。该项目的目标是开发一个数据驱动的物理模型，其解决方案可以重现观测到的飓风丰富而复杂的强度-大小关系。一个成功的项目将有助于更好地理解控制飓风强度-大小关系的物理原理。该项目将提供一个强大的工具，以确定业务预报模型在再现丰富而复杂的强度-大小关系方面的主要缺陷，从而改善预报和公共安全和经济效益。该项目将培养一名大气动力学和数据科学领域的博士后，并通过与两名本科生合作进行荣誉论文研究来支持STEM教育。pi将积极与美国气象学会北佛罗里达分会合作，并向佛罗里达州立大学/塔拉哈西社区提供我们对飓风强度的实验性实时评估。该项目将采用一种新颖的方法，将理论和数据驱动技术结合起来，为飓风的强度-大小关系建立一个新的模型。具体来说，pi将利用数据分析技术来探索控制不同飓风向内径向速度变化的因素，然后将这些因素与飓风角动量损失的变化联系起来。该项目的另一个新颖之处在于，将通过检验其预测方位风径向剖面的能力来验证所开发的模型，该模型既可以从外半径处的边界条件向内预测，也可以从内半径处的边界条件向外预测。预测器和预测器的这种互换性允许人们测试数据驱动的模型是否具有物理定律的质量。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。