Does low accessibility inevitably lead to green deprivation? A study on green space equality from a behavioral...

Green space equality is fundamental to urban environmental justice and equal public service provision. However, existing studies have largely focused on static spatial distributions and supply, with less attention to the dynamic relationship between residents' mobility behavior and green space exposure. From a spatial-behavioral interaction perspective, this study develops an analytical framework for assessing equality in green space exposure. Taking the built-up area of Suqian, China, as a case study, we integrate mobile phone signaling data with multi-source geospatial data, apply the Gaussian two-step floating catchment area (Ga2SFCA) method to measure green space accessibility, and use origin-destination (OD) analysis to characterize long-distance (>1 km) green space visitation at the community level. Among the 319 communities identified as having low green space accessibility, 138 (43.3%) maintained relatively high levels of long-distance visitation, exhibiting a compensatory pattern in which residents accessed alternative green space services through cross-area travel. In contrast, 60 communities (18.8%) simultaneously exhibited low accessibility and low visitation, indicating a potential risk of polarization. These findings reveal marked differences among communities in their capacity to compensate for insufficient local green space provision through travel. Group comparisons showed that compensatory communities generally had higher housing prices, denser transport facilities, and more central locations. Spatial analysis further indicated that both compensatory and polarization communities exhibited spatial clustering, with a more pronounced pattern among polarization communities. Temporal analysis showed that mean hourly visitation between 09:00 and 17:00 was approximately 13.4% higher on weekends than on weekdays. Cross-period comparison further indicated that the median community-level long-distance visitation rate decreased from 0.1099 in May to 0.0676 in December, while community visitation rankings remained significantly and positively correlated across the two periods (Spearman's ρ = 0.403, permutation p = 0.0002), suggesting a degree of temporal persistence in relative visitation differences among communities. The findings demonstrate that low spatial accessibility does not necessarily result in low levels of observed green space visitation, although the capacity for behavioral compensation varies substantially across communities. This study expands the connotation of green space equality from spatial distribution equality to behavioral-based accessibility equality, and provides evidence for identifying risks of green deprivation and developing differentiated urban green space planning strategies.