Air pollution observation—bridging spaceborne to unmanned airborne remote sensing: a systematic review and meta-analysis

Abstract

Air pollution is one of the most critical environmental concerns affecting human health and ecosystem sustainability. This comprehensive review analyzes the evolution and current state of Remote Sensing (RS) methods for air pollution monitoring, examining over 241 relevant papers from the Scopus database using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The study systematically evaluates three main approaches: spaceborne, Manned Aerial Vehicle (MAV)-borne, and Unmanned Aerial Vehicle (UAV)-borne RS. Our analysis reveals significant technological advancements in sensors, platforms, and data processing methods. Spaceborne monitoring demonstrates enhanced spatial resolution (from 10 km to sub-kilometer) and temporal frequency (from monthly to near-real-time). MAV-based systems show superior regional mapping capabilities but face operational constraints. UAVs emerge as promising solutions for local-scale monitoring, particularly in hazardous environments, offering operational flexibility, cost-effectiveness, and the ability to capture high-resolution spatial data. The Internet of Things (IoT) has enhanced data collection networks, while integration of Artificial Intelligence (AI), specifically deep learning, has revolutionized data processing capabilities. Cloud computing platforms, particularly Google Earth Engine (GEE), have further transformed the scale and efficiency of big data analysis for air quality. The meta-analysis of COVID-19 lockdown impacts shows significant pollution reductions, with an overall average decrease of 28% across major pollutants (NO2, PM2.5, PM10, SO2, CO), though individual pollutants showed varying responses, with O3 notably demonstrating increases due to atmospheric chemistry dynamics. The review identifies current limitations and future directions, emphasizing the need for improved multi-platform and multi-sensor RS data integration, sensor miniaturization, and regulatory frameworks. This comprehensive analysis provides valuable insights for researchers, policymakers, and practitioners in environmental monitoring and public health.