Microsoft says it is Aurora Ai that can accurately predict air quality, higher, and more

One of the most recent Microsoft AI models can predict the accuracy of air quality, hurricanes, hurricanes and other weather -related phenomena, as the company claims.

in A paper published in the Journal of Nature Accompanying Blog post This week, Microsoft has detailed Aurora, which says the technology giant can predict the events of the atmosphere accurately and quickly more than traditional meteorological methods. Urra, which is trained in more than a million hours of satellite data, radar and weather stations, simulation, and expectations, can be adjusted with additional data to make specific weather forecasts.

The weather for Amnesty International is not new. Google DeepMind A released handful During the past few years, including WeathernextThe laboratory claims to exceed some of the world’s best prediction systems. Microsoft places Aurora as one of the best performance artists in the field – and a potential grace of laboratories that teach weather science.

In experiments, Ura expected a Dixori hurricane in the Philippines four days before the actual event, overcoming some expert predictions, says Microsoft. The model also outperformed the National Hurricane Center in predicting tropical hurricanes for a period of five days for the 2022-2023 season, and expected the success of the Iraqi Sand Storm 2022.

While Aurora requires a large computing infrastructure for training, Microsoft says the model is very effective in operation. It generates forecasts in seconds compared to the hours that traditional systems use using high computer devices.

The Microsoft, which made the source code and weights available for the public, says it includes AI AI modeling in Aurora in the MSN weather application via a specialized version of the form that produces expectations every hour, including clouds.