Analysis of Surface Level PM2.5 Measured by Low-Cost Sensor and Satellite-Based Column Aerosol Optical Depth (AOD) over Kathmandu

Abstract

A low-cost PurpleAir PA-II sensor was installed, in 2020 at the Institute of Engineering (IOE) Pulchowk Campus, TU located in Kathmandu valley, Nepal, to measure particulate matter with an aerodynamic diameter equal to or smaller than 2.5 µm (PM2.5). The observation shows that hourly averaged PM2.5 fluctuates bimodally in four seasons (Winter: December, January, and February; Spring: March–May; Summer: June–September; and Autumn: October–November), with the highest levels occurring during morning and evening rush hours. PurpleAir records PM2.5 with a maximum average of 101 ± 26.31 µg m–3, in winter, 55.58 ± 11.42 µg m–3, in spring, 45.46 ± 12.16 µg m–3, in autumn, and a minimum of 22.78 ± 3.23 µg m–3, in the summer. Due to rain and diffusion in the vertical atmosphere, PM2.5 levels are lowest during the summer. The ± number for each season represents the standard deviation from the hourly average. AOD550nm data collected by MODIS (Moderate Resolution Imaging Spectroradiometer) onboard two NASA satellites, Terra and Aqua, are compared with simultaneously observed PM2.5. With humidity correction factor f(RH), R2 increases from 0.413 to 0.608 (in winter), 0.426 to 0.508 (in spring), and 0.083 to 0.293 (in autumn). The summer AOD data and PM2.5 are not compared due to a lack of AOD observations. By comparing the column-integrated aerosol data with the surface-level aerosol concentration, this study illustrates the relevance of atmospheric parameters while investigating the reliability of PurpleAir measurements. A cluster analysis of five-day back trajectories of air masses arriving at different altitudes in different seasons indicates that long-range transport of air pollution contributes to MODIS's column integrated AOD by adding aerosol population.