Author: Elisabet De Jesús-Otero
Project Mentors: Audra Luscher and Melida Bailey
Technical Mentors: Lead- Edward Davis; Supporting -Greg Dusek and John Khun
National Oceanic Service (NOS) and National Weather Service (NWS)
National Oceanic and Atmospheric Administration (NOAA)
Rip currents - dangerous currents of water moving away from the coastline - continue to be a leading cause of coastal drownings and rescues, with an annual average of 100 fatalities per year. Despite advances in technology and the accessibility of the National Weather Service (NWS) rip current Beach Hazard Statements, rip current-related deaths have not been prevented. Local Weather Forecast Offices issue these statements during a moderate to high likelihood of rip currents forming at the shore. As of August 12, 2024, 30 fatalities have been reported due to rip currents this season, highlighting the urgent need for more accurate rip current modeling, forecasting and improved risk communication. This project aimed to improve forecast resolution by refining the experimental NOAA Center for Artificial Intelligence and National Ocean Service (NOS) probabilistic rip current forecast model using machine learning techniques. The model analyzed several datasets, such as water level data, wave data and qualitative rip currents lifeguard reports. To verify the model’s accuracy, lifeguards reports for various beaches located around Wilmington, NC were compared with model predictions. Additionally, this project aimed to further improve NWS rip current risk communication and hazard statements by conducting a focus group for various stakeholders, including emergency managers, lifeguards, media, among others. The feedback gathered will inform model development and improve probabilistic rip current forecasts, as well as enhance strategies to better communicate risk and help beachgoers avoid and escape dangerous conditions.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner