Spatiotemporal Pattern of Human Exposure to Air Pollution and the Relationships with Wind-Related Urban Form Factors

Keywords: Human exposure; Strava, Spatiotemporal patterns, Random Forest, Wind-related urban form
Session Type: Virtual Poster
Day: Friday
Session Start / End Time: 4/9/2021 09:35 AM (Pacific Time (US & Canada)) - 4/9/2021 10:50 AM (Pacific Time (US & Canada))
Room: Virtual 52

Abstract

Understanding spatiotemporal patterns of human exposure to air pollution are critically beneficial for outdoor exercisers to examine tradeoffs between physical activities and exposure hazards. However, few studies have focused on the potential relationship between human exposure and the local context, especially for urban form. This study collects mobility data of outdoor exercisers from a crowdsourced database, Strava, analyzes spatiotemporal patterns of human exposure, and further investigates the synergic effects of meteorological factors on urban form-human exposure relationship with a machine learning method, Random Forest (RF). The results reveal several sweet-spots with relatively low levels of pollutant concentrations and high levels of outdoor activities in Atlanta's northern regions. In addition, the wind-related urban form models outperform their counterparts that do not consider the wind factor, demonstrating that incorporating meteorological factors into urban form can improve the understanding of human exposure within urban landscapes. In addition, the study also shed light on that urban form could impact human exposure by affecting the individual preference of outdoor activity locations and corresponding mobility patterns. The results can help decision-makers create a healthy built environment and make air quality management plans to effectively reduce human exposure to air pollutants.