Colorado State University is Developing Artificial Intelligence to Predict Hailstorms
Researchers at Colorado State University are working on a new AI-powered forecasting system to improve the detection of hailstorms and increase warning times. The project aims to extend the current 30-minute warning window to days in advance.
Atmospheric and radar scientists at Colorado State University (CSU) are developing an artificial intelligence-based forecasting system to detect destructive hailstorms earlier. Funded by the National Science Foundation (NSF) through the ASCEND Engine, the study aims to overcome the warning time limitations of current weather models, which are under 30 minutes.
The research team plans to forecast hail risk by combining two different approaches. Developed with NVIDIA, the short-term 'nowcasting' model called 'Hail-StormScope' aims to provide a 2-to-3-hour warning window, while the machine learning model (CSU-MLP) led by Russ Schumacher analyzes nine years of historical weather data to predict storm probabilities within a 1-to-8-day window.
Kristen Rasmussen, associate professor of atmospheric sciences, stated that the project aims to reduce economic risks and insurance costs caused by increasingly frequent and severe hailstorms. The team plans to validate their models with new Vaisala RM60 sensors to be deployed near Greeley, Colorado, in the spring of 2027.
Sources
- thefencepost.com · Link