.webp)
Google’s AI predicts cyclones more accurately, supporting US forecasts amid concerns over funding cuts and data access.
Google introduced a new AI-powered model to help in quicker and more accurate tropical cyclone forecasting. The U.S. National Hurricane Center (NHC) has partnered with Google DeepMind and Google Research to examine the tool that is part of an innovative platform called Weather Lab that shows Google's AI weather analysis.
The platform can forecast a cyclone's route, size, and strength more than fifty times up to 15 days before the date. Google states that in 2023 and 2024, their forecasts were, on average, 87 miles (140 km) nearer to a storm's actual path compared to those of ECMWF, Europe's largest weather service.
This AI system depends on Google's existing GenCast model that exceeded standard forecasting models in 97% of situations. It was developed using data from Europe's ERA5 climate database. Google warns that Weather Lab isn't meant for public forecasting in the present time, instead serving as a research tool.
The project comes following challenges in U.S. weather and climate research. There are fears about the future of free, accessible weather information because of the Trump administration's limitations to organizations like NOAA and the National Meteorological Service. Some fear that if weather services are marketed, accessibility may become further limited. According to experts, although AI systems like these show commitment, they are unable to entirely replace traditional forecasting, which depends on data from weather balloons and satellites.
.webp)
.webp)
