Satellite-Based Analysis on the Associations of Neighborhood Greenness with Demographic and Socioeconomic Characteristics in Seven Korean Metropolitan Cities

Urban greenness is unevenly distributed, but its associations with social and demographic conditions vary across urban settings. Using Landsat imagery and neighborhood statistics, we examined greenness in 1154 neighborhoods across seven Korean metropolitan cities. Spatial models used standardized values of greenness and the continuous predictors, with city fixed effects. In the reference spatial error model, child share was positively associated with greenness (β = 0.250) and log-transformed population density negatively associated (β = −0.593; both p < 0.001). Older-adult share (β = 0.056, p = 0.090) and assessed housing price (β = −0.084, p = 0.052) had fewer certain associations. The National Basic Livelihood Security (NBLS) recipient rate had little marginal association with greenness (r = −0.008) and an uncertain negative estimate when density was omitted. Its jointly adjusted association was small and positive among similarly dense neighborhoods (β = 0.064, p = 0.004), but depended on age-composition adjustment and weakened when large administrative units were excluded. Supplementary spatial screening identified 91 exploratory candidate neighborhoods combining low greenness and elevated vulnerability, with 39 retained after Benjamini–Hochberg adjustment. These candidates support local investigation; their distribution does not establish differences in greenness deficits between cities. The findings support distinguishing deprivation, age composition, and residential concentration when interpreting conditional associations with neighborhood greenness.