High-Resolution Spatiotemporal Analysis of Air Quality and Urban Heat Island in Chicago Using the Microsoft Eclipse Network

Over 100 low-cost air quality sensors were deployed across Chicago from July 2021 to March 2022 during the Microsoft Eclipse project. These sensors provided high temporal and spatial resolution measurements of Particulate Matter (PM_2.5), Ozone (O₃), Nitrogen Dioxide (NO₂), Carbon Monoxide (CO), and temperature. The spatiotemporal variations of the pollutant concentrations reveal distinctive hotspots predominantly located near major expressways and industrial corridors. Hourly PM_2.5 shows strong variation across the sensor network, with standard deviations of approximately 20 µg/m³. The spatial variation decreases with increasing averaging time, and the standard deviations of the daily and monthly PM_2.5 are 5 µg/m³ and 2 µg/m³, respectively. Diel variations of PM_2.5 across the network consistently show a peak around 8-9 am throughout the year, and the peak is attributed to emissions during rush hour. A second peak of PM_2.5 was observed around 8 pm in the fall. Monthly average NO₂ and O₃ mixing ratios exhibit strong seasonal variations with NO₂ reaching the minimum and O₃ peaking in July. Monthly PM_2.5 varies from 8.1 µg/m³ to 13.1 µg/m³ and is elevated in July, November, and December. The seasonal variation of PM_2.5 is likely due to a combination of emission, atmospheric chemistry, and boundary layer development. The sensor network also provided ideal datasets for studying the Urban Heat Island (UHI) in Chicago. Urban heat island intensity (UHII), defined as the average temperature difference between urban and suburban/rural areas, reaches highest value during the summer. The diel and seasonal variations of UHII, and the interactions between UHI and urban pollution island in Chicago will be discussed.