Google says AI weather model masters 15-day forecast

Google Labs says a new artificial intelligence-based weather model can provide 15-day forecasts with unparalleled accuracy and speed, potentially saving lives as climate change intensifies.

GenCast, invented by Google DeepMind, a London-based artificial intelligence research lab, “demonstrated better predictive capabilities” than the current world-leading model, the company said on Wednesday.

The European Center for Medium-Range Weather Forecasts (ECMWF) provides forecasts for 35 countries and is considered the global benchmark for meteorological accuracy.

But DeepMind says GenCast surpasses center accuracy predict Of the 1,320 real-life scenarios in 2019, more than 97% were tested.

Details of its findings Published exist naturea leading scientific journal.

ECMWF president Florence Rabier told AFP the project was a “first step” in integrating artificial intelligence into weather forecasting, but “it’s really a leap forward”.

At this stage, she said, it could be used to complement their current model.

“We are improving year by year,” she added. “In the context of extreme social pressures from climate change, any new approaches that can enhance and accelerate this progress are very welcome.”

extreme weather

The model was trained on forty years of temperatures, wind speed Using air pressure data from 1979 to 2018, it takes just 8 minutes to produce a 15-day forecast, whereas it currently takes several hours.

“GenCast provides better daily weather forecasts and extreme events Up to 15 days ahead of top operating systems…,” DeepMind’s statement said.

DeepMind says GenCast “consistently outperforms” current leading prediction models in predictions extremely hotextremely cold and high wind speeds.

“Predict risks more accurately extreme weather It can help officers protect more lives, avoid damage and save money,” DeepMind said.

Extreme weather is becoming more common and severe due to man-made causes climate change.

In August 2023, a series of wildfires broke out in Hawaii, killing about 100 people. Authorities were criticized by locals who said they had no warning of the impending fire.

This summer, a sudden heat wave in Morocco killed at least 21 people in 24 hours. In September, Hurricane Helene killed 237 people in Florida and other states in the southeastern United States.

“I believe that weather forecasting systems based on artificial intelligence will continue to gradually improve in the future, including better predicting extreme events and their intensity. There is still a lot to be improved in this area,” said David Schultz, a professor at Harvard University. Professor of Weather and Meteorology at the University of Manchester, who was not involved in the study.

But he said these forecast systems rely on already operational weather forecast models, such as those run by ECMWF.

© 2024 AFP