Report: Artificial intelligence has the potential to forecast hurricane landfall earlier

Researchers say that Artificial Intelligence (AI) could potentially save lives by providing earlier warnings about where a hurricane will hit land compared to traditional forecasting systems. Google DeepMind has developed an AI tool called GraphCast, which accurately predicted the landfall of September's hurricane Lee in Canada three days in advance using historical data. While traditional weather forecasting involves complex calculations and takes hours to produce forecasts, GraphCast uses machine learning to quickly digest massive amounts of data and predict weather patterns based on past events. The tool has been successful in predicting severe events like extreme temperatures and tracking the path of storms. However, AI models will complement, not replace, traditional weather forecasts as they still rely on gathering data to train the AI models. The open-source technology is being used by various technology companies and weather organizations worldwide to improve weather forecasting. Although AI shows significant progress, climate change and evolving extreme weather patterns pose challenges for AI-based tools in capturing new weather extremes.