A new weather model based on artificial intelligence can deliver 15-day forecasts with unmatched accuracy and speed - through apps that could save lives as climate change increases, Google said today.
"GenCast," invented by London-based artificial intelligence research lab Google DeepMind, demonstrated better prediction capabilities compared to the current model, the company said today.
The European Centre for Medium-Range Weather Forecasts (ECMWF) provides forecasts for 35 countries and is considered a global benchmark for meteorological accuracy.
But DeepMind said GenCast exceeded prediction accuracy in more than 97% of 1,320 real-world scenarios from 2019. Details of its findings were published in Nature, a leading scientific journal.
ECMWF chief Florence Rabier said the project was a "first step" towards integrating AI into weather forecasting.
"We are making progress year after year. Any new method that can improve and accelerate this progress is extremely welcome in the context of the extreme pressures of climate change," she said.
The model was tested for four decades and can provide a 15-day forecast in just eight minutes – compared to the hours it currently takes.
