Microsoft researchers claimed that its Aurora AI model outperforms operational forecasts in predicting air quality, ocean waves and more. — Pixabay
BERLIN: Artificial intelligence could soon be used to generate faster predictions and more timely warnings of imminent typhoons and downpours, according to scientists based in the US and the Netherlands.
In a paper published by the journal Nature, Microsoft researchers claimed that the company's Aurora AI model "outperforms operational forecasts in predicting air quality, ocean waves, tropical cyclone tracks and high-resolution weather, all at orders of magnitude lower computational cost."
Microsoft stated that the system was "trained on more than one million hours of diverse geophysical data" – enough information to enable it "more accurately predict not just the weather, but a wide range of environmental events in a series of retrospective analyses," including, hurricanes, typhoons and ocean waves
Meanwhile, a team from the US government’s National Oceanic and Atmospheric Administration (NOAA) and the University of Oklahoma said that a prediction model they created using Google DeepMind's Graphcast tool could be 10 times faster at predicting storms than previous computer-based platforms.
The researchers "trained" the Google platform on data from NOAA's Warn-on-Forecast-System (WOFS), creating an AI variant called WOFSCast that shortens forecast times from minutes to seconds.
"The model yielded largely accurate predictions of how storms would evolve for up to two hours; these predictions matched 70% to 80% of those generated by the Warn-on-Forecast system," according to the American Geophysical Union, which published the NOAA/Oklahoma findings its Geophysical Researh Letters journal. – dpa/Tribune News Service
