This paper has presented an analysis of future disparities in heat exposure for different sociodemographic groups from present-day to 2100 utilizing climate model scenarios that provide several possible trajectories of temperature rise, taking into consideration variations in population growth, economic development, and technological change (through the IAM RCP-SSP scenarios) that may impact greenhouse gas emissions and subsequent warming. To our knowledge, this study represents one of the first analyses that quantifies both the spatiotemporal and sociodemographic disparities in HI at the county level throughout the contiguous US. Our findings indicate that Southern counties, with the largest percentage of people of color, specifically non-Hispanic Black communities, encounter the most substantial HI increases over time and exhibit the most significant relative increases when compared to other demographic groups, particularly non-Hispanic White populations. In terms of age, populations over the age of 65 will be disproportionately exposed to increased HI levels, compared to populations below the age of 65, in every scenario examined.
These results underscore the need for understanding differences in underlying sociodemographic factors when evaluating future heat impacts and temperature changes due to climate change. Below we discuss three considerations when applying these results for future studies or policy applications.
The Southern US stands to be most vulnerable to rising heat risks due to sociodemographic factors
Our findings consistently demonstrate that HI in the Southern US remains persistently high across different climate scenarios up until 2100. Although we did not account for migration and that racial, ethnic and age groups may substantially shift from present-day patterns, we note that various SSP scenarios similarly lack substantially distinct demographics (see Fig. S4). The counties with the highest HI increase are collocated with counties with higher percentages of non-Hispanic Black populations and people over the age of 65 across different SSP-RCP scenarios. However, our results still represent a conservative estimation since we use the monthly average HI. Extreme heat days, which can exceed several standard deviations above the monthly average HI, have the potential to pose even greater risks and disparities in a few days 19. Additionally, impacts of heat exposure are highly context-dependent, where local urban design (such as the shade of buildings or trees) or availability of air conditioning play critical roles in individuals’ perception of heat stress 20.
Our estimation shows geographical overlaps between increases in HI and demographics for both race and age, particularly in the Southern US, where the majority of the US non-Hispanic Black population and elderly are located. These demographic groups are documented to be more vulnerable to heat exposure 27,28. Our disparity analysis reveals that HI increases will disproportionally affect non-Hispanic Black populations compared to other racial groups. Furthermore, in the worst-case climate warming scenarios, the gap between HI increases for non-Hispanic Black populations and non-Hispanic White populations widens over time. Other racial and ethnic groups do experience a smaller HI disparity gap compared to non-Hispanic white populations, and this gap narrows in the future in every scenario examined.
Regarding age, our study highlights a growing disparity in heat exposure exposure between individuals aged 65 + and those below, emphasizing the heightened vulnerability of aging populations to rising HI levels. Existing studies have found that “people aged 65 + have been several times more likely to die from heat-related cardiovascular disease than the general population” 29. This finding underscores the critical need for prioritizing climate adaptation and mitigation strategies for elderly people, who stand to be at a greater risk of heat-related health effects due to increased exposure to HI, compared to their younger, healthier counterparts 6,30.
Complexity of projecting Heat Index using Earth System models
Existing studies use different indicators to measure the impacts of increased temperature from climate change. The majority of the studies use remotely sensed land surface temperature 14,15,26,31,32, Heat Indices 21,33, and Wet Bulb Globe Temperature (WBGT). Our analyses reveal substantial disparities in both the magnitude of change and spatial patterns between near-surface air temperature and HI. Our results confirm that global mean air temperature is nonlinearly related to heat stress, meaning that the same future warming experienced could trigger larger increases in societal and health impacts than documented using only surface temperature 14. Our results are also consistent with Chakraborty et al.34, who found differences in the magnitude of disparities in heat exposure and stress when using different metrics, although consistent disparities were determined for populations of color and lower income. The selection of heat indicators should be guided by a study’s specific objectives. When analyzing health impacts, the perception of heat, compared to land surface temperature, becomes a crucial indicator to consider. Additionally, it is worth considering whether utilizing more comprehensive heat indicators can yield more robust and accurate projections when estimating future outcomes.
When utilizing IAMs, it is important to consider various sources of uncertainty in the underlying model assumptions and their projected future scenarios. While measures in standardizing inputs across various IAMs were taken to enhance the comparability of model outputs 35, the difference in outputs is still influenced by parametric uncertainty, model structural uncertainty, and other uncertainties that cannot be fully captured 36. In our analyses, we employ an ensemble of model results to reduce uncertainty across models, although it does not completely eliminate overall uncertainty. Uncertainty also exists in projecting population growth under different SSPs. These SSPs, widely used in scientific assessments like the IPCC Sixth Assessment Report, provide a framework for a set of five climate scenarios (Table S1). However, it is critical to acknowledge that population projections within these scenarios introduce additional uncertainty. For instance, we pair each climate scenario with a specific SSP (Fig. 5), resulting in a range of population that will experience impacts by the 2100s. Projections diverge sharply after 2050, partly due to the incorporation of alternative assumptions regarding key uncertain parameters, particularly at higher levels of population and productivity growth, and equilibrium climate sensitivity 36.
Precisely projecting heat-related impacts and inequality is even more challenging. Most Earth System Models and IAMs are developed at the global scale. National or state-level HI projections are insufficient to analyze demographic disparities between different geographic populations. Migration, which has not been factored into IAMs, represents one crucial factor. While scholars have developed methodologies and techniques to downscale the demographic composition to higher spatial resolutions, achieving precise results (e.g., census block level) is not guaranteed due to increased uncertainties associated with additional assumptions in the downscaling process. Analyzing demographic patterns at higher spatial resolutions, such as census tract or census block level, are impeded by data constraints.
Designing just adaptation measures
The findings of our study highlight a consistent demographic disparity in future heat exposure, revealing greater and growing gaps in exposure for non-Hispanic Black populations and the Elderly, particularly in the Southern US, across all climate scenarios, even with a conservative estimation. Our analyses did not consider the adoption of additional adaptation measures, such as air conditioning, aggressive greening, and assumed no changes in vulnerability (e.g., other underlying health conditions that may become more chronic under climate change). Implementing preventive measures to address these disparities in heat exposure is critical to shaping future climate policies to address what Frosch et al. refer to as the “climate gap,” in which African American and Latino communities already face disproportionate health and economic consequences due to climate and environmental hazards 37. In addition, statistics also show that labor-intensive outdoor industries, such as manufacturing, are primarily located in the South38,39, potentially posing higher health risks to labor forces and economic loss to society 40,41.
Analysis of mid and long-term impacts of climate-related heat exposure on different socioeconomic groups is crucial to develop adaptation plans at the local scale. According to 42 , since just adaptation planning requires the inclusion of socially vulnerable populations, knowing where and whom these populations are is critical to engage them in processes that ensure their involvement in planning decisions that ultimately affect them. With recent US federal policies like the Inflation Reduction Act earmarking specific funds to advance environmental and climate justice, cities and states have access to nearly $3 billion USD to undertake environmental and climate justice activities to benefit underserved commiunities 43 .