Share this story!
- Google DeepMind's AI model forecasts weather in seconds, surpassing traditional methods.
- GraphCast predicts up to 10 days in advance with remarkable accuracy.
- Significant advancement in weather prediction using AI technology.
Google DeepMind's AI model, GraphCast, has introduced a rapid and accurate method for weather forecasting. This AI system can predict weather up to 10 days in advance, utilizing a single desktop computer.
Rapid and precise predictions
Traditionally, weather forecasting is a resource-intensive process, involving high-performance supercomputers that take approximately six hours to generate accurate forecasts. In contrast, GraphCast leverages nearly four decades of historical weather data, allowing it to deliver predictions in less than a minute.
GraphCast's predictive capabilities significantly outperform existing systems, achieving greater accuracy in 90 percent of 1,380 metrics. Its proficiency in forecasting severe weather events, such as extreme temperatures and cyclone trajectories, further emphasizes its superiority. The AI model's energy efficiency is another notable aspect, being about 1,000 times more energy-efficient than traditional methods.