Methods to Estimate Personal Exposure Levels to Air Pollution from Extensive Stationary Air Quality Dataset and Human Mobility Dataset

Abstract

Accurately assessing personal exposure to air pollution has long posed a challenge due to limitations in conventional monitoring approaches. Most studies still rely on sparse, stationary regulatory monitors, assigning identical exposure values to individuals regardless of their movements. This approach neglects the dynamic nature of human mobility patterns and activity locations, leading to inferential errors. This method developed an approach by integrating high-resolution global positioning system (GPS) trajectory data from 100 participants with air quality data from 213 PurpleAir low-cost stationary monitors across Eastern North Carolina. Using geostatistical modelling, which is an automated kriging (ordinary kriging) algorithm developed in Python, the method estimates individualised PM2.5 exposure every minute over a 3-day window (two weekdays and one weekend day), which encompasses 129,600-minute points. The study offers an innovative fusion of spatial and temporal data that bridges the gap between environmental sensing and actual human experience, and the result is a transformative methodology that significantly enhances the precision of personal air pollution exposure assessments from stationary air quality sensors.