Integrating Fixed Monitoring Systems with Low-Cost Sensors to Create High-Resolution Air Quality Maps for the Northern China Plain Region

WashU affiliated authors: Chun-Ying Chao, Huang Zhang, Melanie Hammer, David Kenney, Randall V. Martin, Pratim Biswas (Dept. of Energy, Environmental, and Chemical Engineering)

Abstract: To address and remediate severe particulate matter (PM) pollution in the North China Plain (NCP), many studies have traced pollution sources by using fixed air quality monitoring stations. However, these fixed monitors have high maintenance costs that make it economically infeasible to construct spatially dense networks for air quality measurement. Alternatively, using satellite monitoring systems and a low-cost air quality sensor network can greatly increase the spatiotemporal resolution of the ground-level PM concentration data for a given region. This study comprehensively examines the performance of China’s EPA monitoring stations (CN-EPA), low-cost PM sensor networks and satellite aerosol optical depth (AOD) measurements. The goal is to improve the spatiotemporal resolution of ground-level PM concentration data for Xinxiang, a typical industrial city in the NCP. The inferred results show that low-cost PM sensors demonstrate high linearity with CN-EPA data sets for PM2.5 concentrations with an R2 value of 0.82. The PM2.5 concentration inferred from the AOD retrievals demonstrates a moderate correlation with fixed monitoring stations with an R2 value of 0.53. To evaluate the impact of human activities on air pollution, four traditional Chinese festivals, Chinese New Year, Tomb Sweeping Day, Ghost Festival, and Moon Festival, are chosen to observe the PM distribution in Xinxiang. Heat-maps of the ground-level PM2.5 concentration reveal pollution hotspots in areas of high population density. Cross-validation is employed to evaluate the accuracy of the created pollution maps. The results demonstrate that pollution maps that were interpolated from data measured by CN-EPA data sets have the smallest root mean squared error (RMSE). Finally, our results show that low-cost PM sensor data can be integrated with traditional fixed air quality measurements to provide more detailed information about emission sources on pollution maps in urban and rural areas. 

Citation or DOI: https://doi.org/10.1021/acsearthspacechem.1c00174