CAREER: Understanding Florida Sea-Breeze Convection Using High-Resolution Remote Sensing and Artificial Intelligence

Sea-breeze thunderstorms occur in warm, coastal regions due to the difference in heating of land and water surfaces. While the basic sea-breeze process is well-known, the details of why some days with seemingly similar conditions produce strong, weak, or no storms is less understood. This award provides funding for a combined observational and Interpretable Artificial Intelligence (AI) study of the specific, small-scale atmospheric conditions that promote sea-breeze events in Florida. More reliable forecasting of sea-breeze thunderstorms and their associated hazards of flash flooding and lightning will directly benefit public safety for millions of Florida residents and tourists, while also supporting critical economic sectors such as agriculture, aviation, and outdoor recreation. The project also involves training the next generation of AI-fluent students and conducting public outreach. This award aims to address the hypothesis that the temperature profile - specifically, inversion layer strength and boundary layer temperature - controls the convective depth, precipitation intensity, and lightning flash rate of Florida sea-breeze storms. Further, the researcher will investigate whether an interpretable AI framework is necessary to move beyond statistics and discover the fundamental, non-linear relationships that govern the sea-breeze thunderstorms. The project will make use of operational radar data from the National Weather Service, focused radar data from a 2022 field campaign, satellite-measured lightning data, and environmental data to construct a multi-platform convective database. That database will then be analyzed using traditional and AI tools to elucidate environmental controls and governing dynamics of sea-breeze convection. 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.