Cyclone Structure Characterization and Evolution Forecasting Using Cluster Analysis and Statistical-Dynamical Modeling

使用聚类分析和统计动态模型进行气旋结构表征和演化预测

• 批准号: 0351926
• 负责人: Jenni Evans
• 金额: $ 46.93万
• 依托单位: Pennsylvania State Univ University Park
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2004
• 资助国家: 美国
• 起止时间: 2004-03-15 至 2008-02-29
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0351926&HistoricalAwards=false
• 关键词: Cyclone; Structure; Characterization; Evolution; Forecasting

The PIs seek to gain insights into the dynamics of cyclone structure evolution in three ways: one is the detection and representation of lower-dimensional features by extending the recently developed cyclone phase space diagnostic to all cyclone structure evolutions; the second uses cluster analysis as an aid in defining preferred paths of structural evolution within the cyclone phase space; in the third a regression based statistical dynamical forecast model based on a reduced predictor set will be developed. In addition, some measure of the uncertainty in present operational models for atmospheric cyclones will be sought through a validation of these models based on cyclone phase space diagnostics against their own analyses and a cyclone phase space diagnostic using satellite-derived observations.

PI试图通过三种方式深入了解气旋结构演变的动力学：一是通过将最近开发的气旋相空间诊断扩展到所有气旋结构演变来检测和表示低维特征；第二是使用聚类分析作为辅助确定气旋相空间内结构演变的优选路径；第三是将开发基于回归的基于简化预报器集的统计预报模型。此外，将根据气旋相空间诊断和利用卫星观测对气旋相空间进行诊断，对这些模型进行验证，以求在某种程度上衡量目前大气气旋运行模式的不确定性。