Satellite-informed smart sensor placement framework for near-optimal PM2.5 monitoring in urban areas

Abstract

Effective air quality management depends on ground-based sensor networks that are both accurate and cost-efficient, yet optimal sensor placement is complicated by the interacting influences of road networks, population density, terrain elevation, and resource constraints. This study proposes a novel multi-criteria optimization algorithm that identifies optimal locations and distribution of PM2.5 monitoring sensors by integrating geographical covariates—including roads, population density, and terrain elevation—with satellite observations of surface PM2.5. The Non-dominated Sorting Genetic Algorithm II (NSGA-II) is applied to solve the resulting multi-objective placement problem. The algorithm is validated through a case study in a metropolitan area, demonstrating its ability to identify optimal sensor locations while reducing the number of required sensors and maintaining high estimation accuracy. The study further highlights the value of satellite observations for generating initial PM2.5 estimates that aid sensor placement decisions. Overall, the comprehensive algorithm optimizes air quality monitoring network design, enabling more effective identification of pollution hotspots, more robust assessment of health risks, and better-informed air quality policy and mitigation strategies.

Key Contributions

• Developed a multi-criteria, NSGA-II-based optimization algorithm for PM2.5 sensor network design.
• Integrated satellite-derived surface PM2.5 observations with geographical covariates (roads, population density, terrain elevation) for sensor siting.
• Validated the framework through a metropolitan case study, reducing required sensor count while maintaining high monitoring accuracy.
• Demonstrated the practical value of satellite priors for pollution hotspot identification, health-risk assessment, and air quality policy support.

Recommended citation: Chang-Silva, R., Tariq, S., Kim, S., Moosazadeh, M., Park, S., & Yoo, C. (2024). "Satellite-informed smart sensor placement framework for near-optimal PM2.5 monitoring in urban areas." Environmental Science and Pollution Research. https://doi.org/10.1007/s11356-024-35568-w