Relationships between urban green land cover and human health at different spatial resolutions

Relationships between landscape patterns and ecological processes can vary with changing resolution. Many studies in ecosystem services and human health rely on spatial-dependent data, yet the effects of changes in spatial resolution on the linkages between landscape and human health are underexplored. This study seeks to address the research gap by exploring the relationships of green land cover and pattern metrics at 1 m, 10 m, and 30 m with life expectancy in the City of Baltimore, Maryland, USA. Spearman’s rho correlation and stepwise and hierarchical regression models were applied. Results showed that the effects of resolution change did not emerge for percent green land cover but were evident in other pattern metrics. Multivariate relationships showed that metrics at 1 m explained the most variability of the relationships between green land cover and life expectancy after controlling for potential confounding factors (adjusted R = 0.776, and 0.752 at 10 m and 0.747 at 30 m). Edge density of coarse vegetation was significantly associated with life expectancy at 1 m (adjusted odds ratio [AOR] = 1.012, 95%CI = 1.004–1.024, p < 0.01) and 10 m (AOR = 1.018, 95%CI = 1.009–1.027, p < 0.001) but not at 30 m. Euclidean distance of fine vegetation had a strong positive association with greater life expectancy at 1 m (AOR = 2.067, 95%CI = 1.185–4.072, p < 0.05) but not at 10 m and 30 m. These findings underscore the importance of acknowledging the effects of resolution on the interpretation of landscape-human health relationships and the need for caution when results are used in planning and management decisions.