Key points are not available for this paper at this time.
BACKGROUND: Spatial inequities in greenspace and built environment characteristics contribute to micro-heat islands, where certain urban neighborhoods experience higher temperatures. Historical redlining, introduced by the Home Owners' Loan Corporation (HOLC) in 1938, systematically denied mortgage loans in racially marginalized neighborhoods. Prior research shows that historically redlined neighborhoods exhibit higher land surface temperatures, lower tree canopy cover, and fewer climate adaptations. However, few studies have examined whether the relationship between redlining and urban heat islands varies by neighborhood gentrification. We quantified associations between historical redlining and contemporary micro-heat islands in Philadelphia, PA, and assessed modification by gentrification. We hypothesized that higher HOLC redlining scores would be associated with higher temperatures, but that gentrified tracts would show weaker associations. METHODS: Micro-heat islands in 2018 were measured as census tract radiant temperature deviations from the city mean. HOLC grades were assigned to 2010 census tracts using areal weighting. Gentrification for years 1990-2000, 2000-2010, and 2010-2020, was categorized by census tracts that were "ineligible," "eligible but did not gentrify," "gentrified," or "intensely gentrified" based on socio-demographic changes across three decades. Spatial regression models estimated associations between redlining and radiant temperature deviations, with stratified analyses for gentrification. RESULTS: Among 384 census tracts, 7.8% were non-redlined, 33.6% were moderately redlined, 44.0% were highly redlined, and 15.6% were ungraded. Highly redlined tracts were 0.60 ± 0.20 °F hotter than non-redlined tracts. Gentrification across three decades did not significantly modify associations. CONCLUSION: Historical redlining modestly contributes to contemporary micro-heat islands in Philadelphia, independent of recent gentrification processes.
Edwards et al. (Mon,) studied this question.