100 m PM2.5 mapping from SDGSAT-1 TOA reflectance: Model development and -evaluation

Abstract

Satellite-based fine-resolution (∼100 m) PM_2.5 mapping remains challenging because broadband multispectral imagery struggles to decouple land and atmospheric signals, limiting accurate local emission source detection in regions with sparse ground monitoring networks. Here, we introduce LAD-GAT, a novel deep learning framework for 100 m-resolution PM_2.5 estimation from SDGSAT-1––the first science satellite mission dedicated to the UN Sustainable Development Goals. Specifically, LAD-GAT builds a high-dimensional scene-attribute graph by combining PM_2.5-relevant geographical features, meteorological dynamics, top-of-the-atmosphere (TOA) reflectance, and estimated surface reflectance (SR). A specialized land-atmosphere decoupling (LAD) module is introduced to separate latent aerosol signals from ground surface contributions, and a graph attention network (GAT) models nonlinear associations between in-situ PM_2.5 observations and the input graph structure. In 10-fold cross-validation, LAD-GAT achieved RMSE = 5.042 μg m⁻³ (R² = 0.875) using SDGSAT-1 TOA reflectance, and RMSE = 9.428 μg m⁻³ (R² = 0.862) with Sentinel-2 TOA reflectance. Incorporating daily SR yielded an 8.68 % accuracy gain over TOA reflectance alone and outperformed multi-day composites by 7.27 %, highlighting the benefit of accounting for SR dynamics in fine-scale PM_2.5 mapping. Overall, leveraging the proposed novel LAD-GAT method, SDGSAT-derived PM_2.5 estimates rival those from Sentinel-2, providing fine-scale data to better support SDG 11.6.2 monitoring and targeted air-quality interventions.