Evaluating Air Quality Status in Chicago: Application of Street View Imagery and Urban Climate Sensors

Abstract

This study in Chicago utilizes Microsoft’s Project Eclipse sensors to assess air quality, integrating Google Street View images with conventional land use extraction tools. Principal component analysis (PCA) is employed to evaluate the unique attributes of street view imagery. XGBoost machine learning regression analysis and SHapley Additive exPlanations (SHAP) value calculation are used to explore determinants' impact on air quality. Results show that air pollution levels generally align with city standards, except for critical issues at the Northern intersection of West Belmont Avenue. Regression analysis and SHAP calculations highlight significant differences in the impact of land use on air quality at this intersection compared to random observations. These findings emphasize the need for city agencies to address existing built environment and land use conditions in the North to mitigate potential harm.

Junfeng Jiao Seung Jun Choi Huihai Wang Arya Farahi