AI now predicts 
air pollution

AI is emerging as a powerful tool in the fight against air pollution, a major global health and environmental threat that claims millions of lives each year. While chemical transport and climate-chemistry models have advanced, their accuracy remains limited.

The growth of Earth observation data is creating new opportunities to overcome these challenges. Deep learning (DL), a type of AI that learns from large amounts of data to recognise patterns and make predictions, is helping to develop scalable, interpretable models for air quality prediction.

A recent study finds that DL is revolutionising air quality forecasting by fusing massive, heterogeneous data sources and uncovering patterns invisible to traditional models. The research is led by professor Hongliang Zhang from Fudan University in collaboration with the University of Manchester. The findings were published in Frontiers of Environmental Science & Engineering.

It finds that through multi-sensor data assimilation, DL integrates satellite, ground, and meteorological observations to fill data gaps caused by cloud interference or sparse monitoring networks, generating seamless, high-resolution pollution maps.

However, current models still falter during extreme pollution events, precisely when accurate forecasts matter most. To address this, researchers highlight transfer learning, ensemble prediction, and synthetic event generation as promising methods to boost model resilience.

Equally crucial is the push toward physics-informed neural networks, which embed chemical and physical laws into AI architectures, bridging scientific understanding with computational prediction. The authors advocate for probabilistic and Bayesian approaches to quantify uncertainty, enabling forecasts that predicts what will happen and how confident we can be.

The study suggests that these advances signal a paradigm shift from black-box models to interpretable, physically grounded forecasting frameworks that bring science closer to real-world decision-making.

“Our vision is to make air quality forecasting not just smarter but also more trustworthy,” says Zhang. “By blending physics-based reasoning with the power of DL, we can open the black box of AI and make its decisions explainable.

“This integration allows policymakers and the public to understand why a pollution event may occur and how we can act to prevent it. It’s about turning prediction into prevention – and data into decisions.”

According to the researchers, DL is expected to play an increasingly significant role in environmental forecasting by improving the timeliness and resolution of air quality prediction. The integration of AI with climate-chemistry models is reported to enhance the ability to anticipate pollution trends and their links to climate change.

The study finds that such approaches can support the development of proactive air quality management strategies and inform policy decisions aimed at reducing environmental and public health impacts.