The scientific portal ArsTechnic has published information about the transition to a new era of weather forecasting thanks to the advanced meteorological artificial intelligence model GraphCast, developed by the Google DeepMind team.

According to the report, GraphCast demonstrated significant superiority over traditional forecasting methods, especially in predicting global weather patterns 10 days in advance. The study confirmed that GraphCast outperformed the world’s leading system run by the European Center for Medium-Range Weather Forecasts (ECMWF) by nearly 90 percent in predicting temperature, pressure, wind speed and direction, and humidity at various levels of the atmosphere.

Particularly important is the fact that GraphCast not only has high accuracy, but also provides fast results. The ability to predict hundreds of weather variables around the world 10 days in advance in less than a minute makes this technology unique in the field of weather forecasting.

This new advance in artificial intelligence in meteorology promises to redefine weather forecasting, highlighting the technology’s potential to improve accuracy and efficiency in the field.