A new AI-powered weather forecast model could work with unprecedented accuracy and significantly faster than existing technology.
Created by Google DeepMind – the web giant’s AI-focused lab – Graphcast is set to revolutionize the process of weather forecasting.
GraphCast can predict weather up to 10 days in advance “More accurate and much faster than the industry’s gold-standard weather simulation system – High Resolution Forecast (HRES), produced by the European Center for Medium-Range Weather Forecasts (ECMWF) Is.” Google DeepMind said in a post on Tuesday.
Notably, the tool can also provide early warning of extreme weather events and more accurately predict the movement of cyclones, giving officials and residents more time to prepare for devastating storms, thereby improving the process. Lives can potentially be saved.
When Hurricane Lee struck eastern Canada in September, GraphQL accurately predicted that it would make landfall in Nova Scotia nine days in advance, while traditional forecasts made the same prediction only six days in advance.
The DeepMind team said that GraphQL has been trained on four decades of weather data, enabling it to learn the cause-and-effect relationships behind Earth’s weather systems.
Notably, GraphCast takes less than 60 seconds to create a 10-day forecast, making it much faster than the traditional approach used by HRES, which, according to the team, is a “supercomputer with hundreds of machines.” “It can take hours to calculate in a computer.”
In comparing the two systems, GraphCast gave more accurate forecasts than HRES on over 90% of the 1,380 test variables and forecast lead times.
“When we limited the evaluation to the troposphere, the region of the atmosphere 6 to 20 kilometers high closest to the Earth’s surface where accurate forecasts are most important, our model outperformed HRES on 99.7% of the test variables for future climate, ” the team said.
As weather patterns evolve in Earth’s changing climate, GraphCast will only improve as it provides higher-quality data.
The team is open-sourcing GraphCast’s model code to provide scientists and forecasters access to the technology. This will allow them to tailor it to specific weather events and adapt it to different parts of the world. ECMWF is already testing the tool.
A study published Tuesday by Science takes a more detailed look at GraphQL.
Google DeepMind said, “The use of AI in weather forecasting will benefit billions of people in their everyday lives.” “But our broader research is not just about predicting the weather – it is about understanding the broader patterns of our climate. By developing new tools and accelerating research, we hope AI can empower the global community to tackle our biggest environmental challenges.