Satellite-based Air Quality Monitoring Using Artificial Intelligence: Research Trends and Future Perspectives

Abstract

Advancements in artificial intelligence (AI) and satellite remote sensing have become pivotal in air quality monitoring. This study presents a comprehensive review of 204 papers published between 2015 and 2025, synthesizing key research trends, limitations, and future directions. Based on this review, we identify three principal research approaches: (1) estimating ground-level concentrations from satellite column density products, (2) retrieving these values directly from top-of-atmosphere reflectance, and (3) addressing data gaps in satellite observations to achieve all-sky estimates. Furthermore, we highlight critical challenges identified in previous studies, including nighttime air quality monitoring, uncertainty quantification, consideration of interactions among multiple air pollutants, and the integration of AI techniques for air quality forecasting. Drawing on these findings, we propose targeted research directions to address existing gaps and advance the field. This study enhances the understanding of AI and satellite remote sensing integration, providing a foundation for future advancements in air quality monitoring. The development of a high-resolution, spatiotemporally continuous monitoring framework is expected to improve public health protection and strengthen climate change mitigation efforts. © 2025 Elsevier B.V., All rights reserved.