The role of gentrification in occurrence of extreme temperature events
On this page:
Methods: Four categories of gentrification for U.S. census tracts (ineligible for gentrification, eligible but not gentrified, gentrified, intensely gentrified) were calculated using a modified protocol from Drexel University Urban Health Collaborative for 1990-2000. Average daily temperature from 2002-2019 for each tract came from the U.S. EPA’s EQUATES model; days above and below percentile-based thresholds were identified for each census tract independently using all years of data. Based on percentile-based thresholds, ECEs were defined as mean daily temperature <2.5th or <5th percentile for 2 or 3 consecutive days. Based on percentile-based thresholds, EHEs were defined as mean daily temperature >95th or >97.5th percentile for 2 or 3 consecutive days. Linear regression models were used to estimate change in temperature variables over the study period by gentrification status (eligible for gentrification as reference), adjusted for beginning-year (1990 or 2000) population density. Results: We observed shifts in temperature distribution for gentrified and intensely gentrified census tracts; gentrification was associated with higher low temperatures and lower high temperatures. Depending on the metric, there were also shifts in EHEs and ECEs. Conclusions: Preliminary analyses indicate that the distribution of temperatures may vary across gentrified and non-gentrified census tracts. Specifically, we observe a compression of the distribution of temperatures (i.e., lower maximum temperatures and higher minimum temperatures) in gentrified compared to eligible, but not gentrified, census tracts.