Associations between Elderly Residents' Supermarket Accessibility and Built Environmental Features in Seoul, Korea

Abstract

Scholars have examined the relationship between destination accessibility and various pedestrian-friendly environments in neighborhoods. However, there have been few studies on whether micro-scale environmental features promote access to supermarkets. This study fills these gaps by examining the accessibility of supermarkets and the meso- and micro-scale environmental features. In particular, environmental features that pedestrians can experience at the eye-level were estimated using the Naver pedestrian direction API and semantic segmentation technique. Multivariate regression models were employed to determine whether neighborhoods with high supermarket accessibility ensure walkable environments for the elderly. Our findings have shown that the walkability differs according to different levels of supermarket accessibility. Neighborhoods with high supermarket accessibility had high density and high diversity, and they provided high-quality sidewalk pavement. Conversely, elderly residents in neighborhoods with low accessibility had to pass through more pedestrian crossing facilities and steep slopes to go to do their shopping. These results can provide insight to policymakers and planners about the uneven accessibility of supermarkets and help improve neighborhood walkability for the elderly.