CAREER: Geospatial Modeling of Tropical Cyclones to Improve the Understanding of Rainfall Patterns

在热带气旋登陆时，大气和地球表面都会改变其结构。这些结构变化影响了风暴产生的降雨的整体空间格局，进而控制了可能发生的洪水的大小和时间。尽管热带气旋模型已经纳入了诸如风切变和地形等地球物理变量来改进对总降雨量的预测，但它们仍然缺乏准确描述降雨量地理分布的能力。该项目解决了改进热带气旋雨场空间模拟的迫切需要。地理信息系统（GIS）中使用的形状分析技术将量化热带气旋产生降雨区域的范围和位置，包括由雷达反射率返回值定义的强降雨区域。然后，通过将观测到的模式与模型生成的风暴数据进行比较，将这些区域与影响风暴结构的关键地球物理变量联系起来。这项由CAREER奖支持的研究将开发一套指标来描述热带气旋雨场经常相似的形状，以确定产生这些形状的过程，并将推导出预测这些形状的统计模型。通过在GIS中对雨场形状进行建模，这些形状可以潜在地合并到现有的基于GIS的水文模型中，以改善这些模型中的降雨量输入。多元逻辑回归分析还将用于预测哪些地区将不下雨，哪些地区将下雨，以及降雨率低与降雨率高。这项研究将通过量化热带气旋雨场和提高确定产生洪水的降雨位置的能力，有助于改进降雨预报。这个项目的一个组成部分是努力教育新一代的研究人员在先进的空间分析技术。这些努力将有助于培养创新的观点，并促进进一步改进对登陆热带气旋的降雨预报。这些成果将通过强调基于gis的气象和气候数据分析的教育计划来实现。研究生和本科生都将接受基于地理信息系统的大气数据分析方面的加强培训，并将开设两门新设计的课程，以便开设气象学本科辅修课程。受资助和未受资助的学生将承担与天气事件空间分析相关的个人研究项目，与来自校外的一系列主题相关的特别客座讲师进行互动，并参加以GIS或天气相关研究为主题的专业会议和研讨会。指导活动预计将增加少数民族学生选择科学为基础的职业的人数。

Both the atmosphere and the earth's surface act to change the structure of a tropical cyclone during landfall. These structural changes affect the overall spatial patterns of rainfall produced by the storm, which then control the magnitude and timing of flooding that may occur. Although models of tropical cyclones have incorporated geophysical variables such as wind shear and terrain to improve the prediction of rainfall totals, they are still lacking in their ability to accurately depict the geographical distribution of the rainfall. This project addresses the critical need to improve the spatial modeling of tropical cyclone rain fields. Shape analysis techniques employed within a Geographic Information System (GIS) will quantify the extent and locations of rain-producing regions of tropical cyclones, including heavy rainfall regions as defined by radar reflectivity returns. These regions will then be linked to key geophysical variables that influence the storm's structure through comparisons of observed patterns against data from model-generated storms. The research supported through this CAREER award will develop a set of metrics to describe shapes that tropical cyclone rain fields frequently resemble to determine the processes responsible for creating those shapes, and it will derive statistical models to predict those shapes. By modeling rain field shapes within a GIS, these shapes can potentially be incorporated into pre-existing GIS-based hydrological models to improve the rainfall inputs into these models. Multivariate logistic regression analysis will also be employed to predict which areas will receive no rain versus some rain, and low rainfall rates versus high rainfall rates. This research will contribute to improving rainfall forecasting, both through the quantification of tropical cyclone rain fields and the improved ability to identify the location of flood-producing rainfall. An integral component of this project is an effort to educate a new generation of researchers in advanced spatial analysis techniques. These efforts will contribute to cultivating innovative perspectives and fostering further improvement of rain forecasts for tropical cyclones that reach land. These outcomes will be achieved through an educational program that emphasizes GIS-based analysis of meteorological and climatological data. Both graduate and undergraduate students will receive enhanced training in GIS-based atmospheric data analysis and two newly designed courses will allow an undergraduate minor in meteorology to be developed. Students, both funded and non-funded, will undertake individual research projects related to the spatial analysis of weather events, interact with a series of topically relevant special guest lecturers from outside the university, and travel to professional meetings and workshops featuring GIS or weather-related research. Mentoring activities are expected to increase the number of minority students selecting science-based careers.