A deep learning model for predicting daily PM<sub>2.5</sub> concentration in response to emission reduction.
basic_science · Level V
Where this comes from
- Record sourced from PubMed, PMID 42467793.
- Also identified by DOI 10.1126/sciadv.aef5759.
- No licence information is recorded for this record.
- Because redistribution is not established, this page shows the abstract only. Follow the links below for the full text.
Abstract
Air pollution remains a leading global health threat, with fine particulate matters (PM<sub>2.5</sub>) causing millions of premature deaths annually. Chemical transport models (CTMs) are essential for estimating how emission controls improve air quality but are computationally intensive. Here, we present CleanAir, a deep learning model that simulates daily PM<sub>2.5</sub> concentration and its chemical composition in response to emission reductions at a 36-kilometer horizontal resolution. Built on a residual symmetric three-dimensional U-Net architecture, CleanAir can estimate 365-day PM<sub>2.5</sub> concentration over China within 10 seconds on a graphics processing unit or 160 seconds on a central processing unit-three to four orders of magnitude faster than CTMs. Results from CleanAir agree well with those from a Community Multiscale Air Quality (CMAQ) model for both PM<sub>2.5</sub> concentration and emission-induced changes. Trained on 2416 emission scenarios from the CMAQ model, CleanAir generalizes well across unseen meteorology and emissions. With fast simulation capability, CleanAir enables extensive evaluation for short-term emission control measures and long-term mitigation pathways, leading to more responsive decision-making.