WeatherIA
FRGBMA
ia-meteo

Aerosols and AI: Neural networks to improve weather forecasts and air quality alerts

Researchers are using machine learning and satellite data to integrate aerosols into predictive models, promising to reduce forecast uncertainty and provide more accurate air quality alerts in France.

WE

Rédaction Weather IA

vendredi 2 octobre 2026 à 19:098 min
Partager :Twitter/XFacebookWhatsApp
Aerosols and AI: Neural networks to improve weather forecasts and air quality alerts

Air quality forecasts remain largely uncertain today: weather models struggle to incorporate the effects of millions of suspended particles, aerosols, which modify cloud formation and pollutant dispersion. An international team has published a study showing how machine learning, coupled with Copernicus satellite data, can reduce forecast uncertainty by several hours, opening the way for more reliable alerts for urban residents and farmers.

What the study reveals: Role of aerosols in predictive models

The analysis shows that aerosols - whether forest smoke, pollen, desert dust or volcanic sulfates - directly influence solar radiation and cloud microphysics. By integrating these particles into predictive models, researchers have observed a 15% improvement in 24-hour temperature forecast accuracy and a 20% reduction in the error of PM2.5 concentrations at 48 hours. These gains are mainly due to better representation of light absorption and diffusion processes, two crucial processes for calculating atmospheric energy balance.

The results come from a comparison between the ECMWF model, which traditionally uses very simplified aerosol fields, and a new hybrid model combining the GraphCast system and the Pangu-Weather neural network. The latter exploits Copernicus atmospheric data enriched by lidar and satellite spectrometry observations to create a high-resolution aerosol field. The study highlights that adding detailed information about aerosols reduces forecast uncertainty, especially during extreme events such as forest fires or sandstorms.

How researchers integrate aerosols: Method and data

The core of the approach is machine learning. Scientists have trained a deep neural network using several years of atmospheric data, including ground measurements, radiosonde profiles, and Sentinel-5P multispectral images. This machine learning process has allowed complex correlations to be established between aerosol concentrations and meteorological variables such as relative humidity, surface temperature, and wind speed. The predictive model thus created is capable of generating aerosol fields in real-time at a 10 km scale, much finer than the classic 50 km grids used by traditional models.

To ensure the robustness of the system, teams have coupled the outputs of the neural network with the ECMWF prediction model and the Copernicus data system. This synergy creates a continuous flow where aerosol forecasts feed atmospheric dynamics calculations,

Was this article helpful?

Commentaires

Connectez-vous pour laisser un commentaire