Sea level rise (SLR) poses a significant threat to coastal communities, causing widespread flooding and potentially isolating populations from essential services due to inundated transportation networks. While inundation itself is a major concern, the resulting isolation from critical services like hospitals and schools presents a less visible but equally serious challenge, particularly for vulnerable populations. This study focuses on understanding who is most at risk of isolation due to SLR, a crucial step in developing equitable adaptation plans. Disadvantaged groups, including racial minorities, older populations, low-income individuals, and renters, are disproportionately affected by climate change impacts. An estimated 20 million coastal residents in the U.S. will be at risk of inundation by 2030. Current literature mainly focuses on direct inundation risk, neglecting the more complex issue of isolation. This research addresses this gap by examining how isolation interacts with social vulnerability to potentially exacerbate existing inequalities. Unequal risk from SLR manifests differently across scales. For instance, in New York City, factors like uneven exposure, inequitable adaptation responses, and historically discriminatory practices contribute to disproportionate vulnerabilities. Underinvestment in infrastructure in marginalized communities further exacerbates their resilience to climate change impacts. This study investigates how the risk of isolation varies among different racial and ethnic groups in coastal U.S. areas, and how this risk correlates with social vulnerability factors like age, income, renter status, and racial composition. The research uses a novel methodology that combines OpenStreetMap (OSM) road network data with National Oceanographic and Atmospheric Administration (NOAA) sea level rise scenarios to calculate isolation risk. A census block is considered at risk if there is no available (non-inundated) route to essential services.

Existing research predominantly focuses on the direct impacts of sea-level rise, such as inundation, neglecting the significant consequences of community isolation from essential services. Studies have documented the disproportionate vulnerability of marginalized communities to climate change impacts. In New York City, for example, studies highlight the intersection of uneven exposure, inequitable adaptation measures, and historical discriminatory practices in exacerbating flood risks among racial and ethnic groups. Underinvestment in infrastructure within historically marginalized communities is also a recurring theme, resulting in a built environment less resilient to natural hazards. The economic considerations of adaptation further highlight the potential for abandonment of socially disadvantaged communities. This study builds upon these previous works by explicitly examining the risk of isolation as a critical dimension of SLR impacts, and by focusing on the interplay between isolation and existing social vulnerabilities.

This study uses a multi-faceted approach to analyze the risk of isolation due to sea-level rise. Firstly, an isolation metric is developed by combining OpenStreetMap (OSM) road network data with National Oceanic and Atmospheric Administration (NOAA) sea-level rise scenarios (1 to 10 feet) for all coastal counties in the continental U.S. Relative sea-level rise (RSLR) is calculated using tidal gauge data and SLR projections to account for non-uniform sea-level rise. A census block is classified as at risk of isolation if its centroid lacks an un-inundated route to essential services such as fire stations or primary schools. These services act as proxies for other critical service areas. The risk of isolation is computed at the census block level, using population data from the 2020 U.S. Census and socioeconomic and demographic data from the American Community Survey (ACS). A distributive justice framework is adopted to identify disparities. The study focuses on Black and Hispanic populations, comparing their proportion in the at-risk population to their proportion in the overall population. The analysis involves aggregating census block-level results at different geographical scales: national, state, and county levels. This allows for a comparative analysis of disproportionate risks across various geographic areas. Census tract level data is also used to assess the interactions between racial minority status and other socio-demographic factors such as income, age, and renter status, to identify compounding vulnerabilities. Generalized Linear Models (GLM) are employed to analyze the relationships between socio-demographic characteristics and both isolation and inundation risks. Three distinct outcome variables are used: (1) percentage of the population at risk of isolation; (2) percentage of the population at risk of inundation; and (3) percentage of the population “missed” in an analysis relying solely on inundation. The GLM models examine various socioeconomic factors and their correlation with the risk of isolation, inundation, and the 'missed' population. The study also considers the temporal aspect of risk, analyzing the timing of disproportionate burdens on racial minorities based on both intermediate and high SLR scenarios.

The study reveals significant racial disparities in the risk of isolation due to SLR. Hispanic populations are overrepresented in the total population at risk of isolation starting at 4 ft of SLR, while Black populations are overrepresented at approximately 6 ft. White populations are underrepresented after 5 ft of SLR. The timing of these disparities depends heavily on the SLR scenario pathway. Under an intermediate scenario (1.0 m global SLR by 2100), strong evidence of disparate isolation effects emerges by 2120, while under a high scenario (2.0 m global SLR by 2100), these effects appear as early as 2090. At the state level, Black populations in Pennsylvania experience disproportionate risk at 3 ft of SLR, while Hispanic populations in Florida, Louisiana, Mississippi, and Maine face similar risks. At the county level, the number of coastal counties with disproportionately at-risk Black and Hispanic populations increases with rising SLR levels, with Hispanic populations showing greater vulnerability. Black populations are disproportionately affected primarily in the Northeast and California, while Hispanic populations are vulnerable across the study area, particularly in Northern California, Louisiana, and parts of the Northeast. Thirty-four counties face disproportionate risk for both Black and Hispanic populations at 3 ft of SLR. At the census tract level, the disproportionate risk of isolation for racial minorities shifts as SLR increases, with some tracts showing isolation rates more than ten times greater than the group’s representation in the overall county population. Analysis of socio-demographic characteristics shows that tracts with disproportionately high isolation risk for Black and Hispanic populations also have high median household incomes, low percentages of White populations, and older populations (especially at higher SLR levels). Generalized Linear Models (GLM) reveal that older age, higher income, higher rates of minority populations, and higher proportions of renting populations are associated with higher percentages of populations at risk of isolation. The GLM also indicates that younger populations, those with higher incomes, more renters, and lower percentages of Black and Hispanic populations are more likely to be “missed” (i.e., not captured) by analyses focused solely on inundation risk.

The findings of this research highlight the significant and disproportionate risks of isolation facing marginalized communities due to sea-level rise. The study's results underscore the importance of considering isolation risk alongside direct inundation risk in climate adaptation planning. The significant number of counties and census tracts with disproportionately high isolation risk for Black and Hispanic populations reveals a stark inequity in the burden of SLR. The differing timing of these effects under various SLR scenarios emphasizes the crucial role of mitigation efforts in reducing the magnitude and delaying the onset of these adverse impacts. The association between isolation risk and socio-demographic factors such as age and renter status points to the necessity of considering these vulnerabilities in adaptation strategies. The finding that populations with higher incomes and younger ages are more likely to be missed by inundation-focused analyses highlights the need for a more comprehensive approach to risk assessment. The study's findings provide strong evidence for incorporating measures of isolation risk within an equity framework into climate adaptation planning and policies.

This study demonstrates that sea-level rise will disproportionately isolate disadvantaged populations in the U.S., particularly Black and Hispanic communities, older adults, and renters. It is crucial for decision-makers to integrate measures of isolation risk, using an explicit equity framework, into adaptation planning and policies. The risk of isolation, and associated uncertainties, can inform dynamic adaptive pathways planning and long-range urban planning decisions. Future research should explore the complex interactions between SLR, isolation, economic investment, and gentrification, as well as finer-scale spatial variations in isolation risk.

The study's analysis considers the entire population of a census block to be isolated based on transportation routes from the block's centroid, potentially missing finer-scale inequities. The analysis does not account for potential future changes to road networks or other adaptation measures. Census data limitations may affect the accuracy of demographic information and risk estimates. Finally, the study does not explicitly consider future demographic shifts, which may influence the spatial distribution of vulnerability.