“AI researcher Paris Perdikaris at Microsoft Research AI for Science in Amsterdam and his colleagues found that Aurora could in less than a minute predict the levels of six major air pollutants worldwide: carbon monoxide, nitrogen oxide, nitrogen dioxide, sulfur dioxide, ozone and particulate matter. Its predictions span five days. It can do it ‘at orders of magnitude smaller computational cost’ than a conventional model used by the Copernicus Atmosphere Monitoring Service at the ECMWF, which predicts global air-pollution levels, the team wrote in a preprint1 published on arXiv on 20 May.

Aurora’s predictions were of a similar quality to those of the conventional model.”

From Nature.