The Uncertain Geographic Context Problem (UGCoP) in Measuring People’s Exposure to Green Space Using the Integrated 3S Approach

Keywords: Urban green space, exposure measure, GPS, GIS, remote sensing, geographic context, mobility, spatial resolution, buffer zone, UGCoP
Abstract Type: Virtual Paper Abstract

Abstract

The “mobility turn” in environmental health studies promoted the integrated 3S (GIS, GPS, and remote sensing) approach in the studies of urban residents’ exposure to green space and corresponding health outcomes. However, few studies have examined the possible errors in measuring exposure to green space using the conventional and integrated 3S approaches. In this paper, we compared the differences in green space exposure obtained from different geographic contexts using residence-based and mobility-based methods and multiple spatial resolutions and buffer zones. We collected the 7-day GPS trajectories from 208 participants at 1-min temporal resolution in Hong Kong. Entire Hong Kong’s green space was delineated using multiple remote sensing data sources at 3 m, 10 m, and 30 m spatial resolutions, respectively. Lastly, the residence-based and mobility-based exposures to green space were calculated for each participant using 100 m, 300 m, and 500 m buffer zones at three spatial resolutions, respectively. Descriptive analyses, paired sample t-tests, and logistic regressions were employed to test the difference between algorithmic settings and their influences on interpreting green space’s health effects. The results indicate multiple significant disparities. The most striking difference is that mobility-based exposures to green space are significantly higher than residence-based exposures, manifesting the uncertain geographic context problem (UGCoP). As a result of this study, we suggest using mobility-based measures of exposure to green space using a smaller buffer zone and finer spatial resolution, which enables more accurate measures of exposure to green space for the study of green space’s health effects.