Jaeda Woodruff ’25
Real-time Forecasting System for Hurricane Hazards and Risk
I compared the performance of various hurricane forecasting models from 2020-2022 and created a customizable forecasting tool trained on publicly available data. The tool uses a super ensemble approach, combining multiple independent models in a performance-based weighted average. This tool will be used to enable real-time, highly accurate hazard forecasting of factors such as wind speed, storm surge, rainfall and storm path on a county-by-county basis without reliance on subjective forecasts. Through this research experience, I learned a range of skills including data analysis in the program Python and how to use simple machine-learning models to minimize error with multiple linear regression. I also gained insights into the strengths and weaknesses of our current forecasting abilities and the unusually active 2020 and 2023 hurricane seasons. I plan to extend my work with extreme weather over the next semester by researching the historical relationship between greenhouse gases and extreme weather formation.
Oceans and Atmosphere
Hurricane Hazards and Risk Analysis Group, Department of Civil and Environmental Engineering, Princeton University - Princeton, New Jersey
Ning Lin, Professor of Civil and Environmental Engineering; Christine Blackshaw, Ph.D. candidate, Civil and Environmental Engineering; Avantika Gori, Ph.D. candidate, Civil and Environmental Engineering