HailStorm is a research project that uses artificial intelligence to better understand where and how often hailstorms occur. By analyzing historical weather data together with reported hail events, the project produces maps that show hailstorm risk across large regions in Europe and the United States. These maps help stakeholders better understand and manage one of the most damaging types of extreme weather.

Problem

Hailstorms cause significant damage every year, especially to agriculture, energy infrastructure and insured assets. Despite their high impact, hailstorms are difficult to quantify and predict. They are caused by complicated atmospheric processes, and reliable observations are frequently rare or unevenly distributed, especially across wide geographic areas. The lack of consistent data makes it difficult to estimate risk and create effective mitigation strategies.

Solution

HailStorm addresses this challenge by using machine learning to identify patterns in past weather conditions and reported hail events from Europe and the United States. The AI model is carefully designed, trained and calibrated to recognize the meteorological conditions that cause hail production. Users can select a region and a time period of interest, and the system automatically generates maps and visual summaries showing where and when hailstorms are most likely to occur. This method allows for the creation of consistent and comparable hail risk maps, even in areas with limited observations, thus facilitating better planning and decision-making for agriculture, insurance, and public agencies.