Examining the neighborhood effect averaging problem (NEAP) and spatial non-stationarity in green space exposure and distribution in the United States

Abstract

Green space is essential for healthy and functional communities, yet access to green space across the United States remains uneven. Research on green space exposure has produced inconsistent findings, partly due to a reliance on static, residence-based measures that fail to fully capture individuals' daily mobility. In particular, limited work has examined how the neighborhood effect averaging problem (NEAP) and spatial non-stationarity influence exposure estimates at a national scale, creating gaps in understanding environmental inequality. This study addresses these gaps by comparing home-based and mobility-based green space exposure across the continental United States. Census block-level commute data are integrated with two green space proxies, WorldCover land cover and a USA Parks dataset. Exposure is analyzed at national and state levels and across income levels, with statistical testing used to evaluate differences and spatial variability. Results show that exposure estimates are highly sensitive to both mobility and dataset selection. Mobility-based measures generally reduce average exposure and compress variability, providing strong evidence of the NEAP, though its magnitude varies geographically. Differences across income groups are statistically significant but small and spatially inconsistent, suggesting income is a weak predictor relative to local context. Additionally, large discrepancies between datasets demonstrate that how green space is defined strongly influences outcomes. These findings highlight the need for mobility-informed, spatially explicit approaches to better capture environmental exposure. Improving measurement frameworks can support a more accurate understanding of environmental inequality and inform public health and urban planning decisions.