More extremely hot days, more heat exposure and fewer cooling options for people of color in Connecticut, U.S.

Urban heat exposure, exacerbated by climate change and urban development, poses significant health, economic, and environmental risks. In the United States, heat is a leading cause of weather-related fatalities and contributes to numerous non-fatal illnesses and economic losses. Existing research consistently shows that people of color (POC) in the U.S. disproportionately experience higher heat exposure, vulnerability, and lower adaptive capacity compared to white populations. However, most prior studies have focused on large cities at single points in time, leaving the temporal dynamics of heat inequality in smaller and mid-sized cities less understood. This study addresses this gap by examining the changing nature of urban heat exposure and adaptation strategies in ten of Connecticut's largest cities, which are representative of mid-sized and small cities in the U.S. The study focuses on two aspects: heat exposure (overall temperatures, extremely hot days, and changes in exposure over time) and adaptation strategies (air conditioning ownership and tree canopy cover). The research questions explored include disparities in overall temperatures, number of extremely hot days, change in heat exposure over time, and differences in tree cover and air conditioning ownership rates between POC and white communities.

The literature extensively documents that people of color in the U.S. face disproportionate exposure to extreme urban heat. Studies reveal that POC communities often experience stronger surface urban heat island effects and higher moist heat stress than white communities. This disparity is further compounded by fewer heat adaptation strategies among POC populations. For instance, Black Americans are significantly more likely to reside in areas with high heat-induced mortality. However, a significant limitation of previous research is its focus on large cities and analysis limited to a single point in time. While small and mid-sized cities comprise a significant portion of the U.S. population and may have high proportions of POC residents, these urban areas have been underrepresented in past research. This study is unique in its longitudinal analysis of heat inequality, examining changes in heat exposure and adaptation options across multiple years in a range of cities, all within a defined study area.

The study area encompasses the ten most populous cities in Connecticut. Data sources included US Census data (1990, 2000, 2010, and 2020) providing racial demographics at the census block level; a global daily near-surface air temperature dataset (2003-2020) at 1km resolution; Landsat Collection 2 land surface temperature (LST) data (1990-2022) for time series analysis; high-resolution (60cm) airborne imagery (NAIP) from 2018 for tree cover mapping; building-level property assessment data (CAMA) from 2022 to determine air conditioning (AC) ownership rates; and OpenStreetMap data. Race data was used to categorize census blocks as predominantly POC or predominantly white. Mean daily maximum summer air temperature was calculated at the census tract level. The Continuous Change Detection and Classification (CCDC) model was employed with LST and NDVI data on Google Earth Engine to analyze temporal trends. AC ownership rates were calculated at the block level using building footprints and parcel-level AC availability. Tree cover and vegetation cover (using NDVI) were mapped and analyzed at the block level using object-based classification of NAIP imagery. Disparity analyses were conducted at both census tract and census block levels, examining differences in temperatures, number of extremely hot days, changes in exposure over time, and differences in tree cover and AC ownership rates. Spatial statistical methods, such as spatial lag models, were also used to analyze relationships between temperature and socio-economic variables. Correlation coefficients were calculated to examine relationships between various variables of interest.

The study revealed significant inequalities in heat exposure and adaptation options between POC and white communities. Predominantly POC communities consistently experienced higher mean summer air temperatures (0.35 °C higher) and land surface temperatures (4 °C higher) compared to predominantly white communities. POC communities experienced significantly more extremely hot days (35 more days on average from 2003 to 2020) and a larger increase in heat exposure per capita over time. This increase was attributed not only to rising temperatures but also to a growing POC population concentrated in hotter areas. Furthermore, POC communities showed significantly lower AC ownership rates (23% lower on average) and substantially less tree canopy cover (15% lower on average). This disparity in AC ownership and tree cover was consistent across all ten cities. The analysis of LST and NDVI over time (1990-2020) confirmed the persistence of these inequalities and demonstrated the limited capacity of vegetation cover increases to offset rising temperatures. The relationship between LST and the percentage of various racial groups highlighted the particularly strong association between LST and the percentage of Black or African American residents.

The findings strongly support the hypothesis that POC communities experience disproportionate heat exposure and have limited access to effective cooling strategies. The persistent disparities across multiple years, despite increased diversity in some cities, underscore the systemic nature of this inequality. The combined effects of higher temperatures and lower adaptive capacity—reduced AC access and tree cover—create a significantly higher risk of heat-related illness and mortality in POC communities. These findings have important implications for urban planning and policymaking, highlighting the need for targeted interventions to address environmental injustice and mitigate heat-related health risks in vulnerable communities.

This research provides compelling evidence of heat inequality in Connecticut’s mid-sized cities, demonstrating that POC communities experience greater heat exposure and possess fewer cooling options. The study underscores the urgency of implementing equitable heat mitigation and adaptation strategies, particularly prioritizing tree planting in underserved communities with low AC access. Future research should expand geographically, include further analysis of indoor cooling strategies, and consider the influence of mobility and working environments on heat exposure. Continued monitoring of heat exposure and access to adaptation options at a fine spatial scale is critical for effective policy interventions and equitable resource allocation.

The study has several limitations. First, the analysis focuses primarily on AC ownership as a measure of adaptation capacity, neglecting other potential cooling strategies. Second, AC ownership does not equate to AC usage; factors like financial constraints might influence usage patterns. Third, the AC data used is based on property assessment and might underestimate window AC units, particularly in rental properties. Fourth, the study does not explicitly account for the mobility of residents, which could affect their overall heat exposure. Finally, this study is limited to the ten largest cities in Connecticut; broader geographic scope is needed to generalize findings to other regions.