© Chris Yarzab/Climate Visuals

We are pleased to announce the publication of a new article, “Experience-driven perceptions misalign with assessed heat risk in the United States,” in Nature Communications. An interactive map illustrates the findings.

Key Findings

• Most Americans underestimate their local heat risk: In the vast majority of U.S. counties, public concern about extreme heat is lower than public-health assessments indicate.
• Experience is a poor guide to future heat danger. People rely on past weather and lived experience, but climate change is pushing heat risk beyond what many communities recognize.
• The biggest danger zones are where risk is high and concern is low. Many rural, older, and higher-poverty counties face serious heat risk with little public awareness—undermining preparedness and adaptation.

These findings align with the experiences of many people who work regularly with and in frontline communities. Thus, it is important to heed the views of people and practitioners in frontline communities not only in decision-making but in the science that informs it.

Extreme heat is the deadliest weather-related hazard in the United States, yet most heat-related illness and death is preventable with appropriate preparation and adaptation. A core challenge, however, is that people’s perceptions of local heat risk often diverge sharply from expert assessments, limiting protective behaviors and effective policy responses.

In this study, we introduce the Risk Analysis–Perception (RAP) framework, which directly compares assessed heat risk (measured using the CDC’s Heat and Health Index) with perceived heat risk, modeled from five years of nationally representative survey data (2018–2022; N = 11,113). Using multilevel regression with poststratification (MRP), we estimate the percentage of residents worried about extreme heat in every U.S. state and county, and then map where perceptions align—or fail to align—with assessed heat risk to human health.

We find large perception gaps across much of the country. In most U.S. counties, assessed heat risk is substantially higher than perceived risk, particularly in the Pacific Northwest, Appalachia, parts of Michigan, and rural areas. Only about 4% of counties have a majority of residents who are at least moderately worried about extreme heat, despite growing health risks as the climate warms. In contrast, a small number of counties—mainly in parts of Texas and California—show high perceived and high assessed risk, indicating stronger alignment between people’s lived experience and their assessed vulnerability.

Demographic and socioeconomic factors play a major role in shaping these gaps. Counties with higher poverty rates and older populations are more likely to underestimate heat risk relative to assessments, while counties with higher levels of education tend to show closer alignment with, or greater awareness of, assessed risk levels. We also find important racial and ethnic patterns that reflect underlying structural vulnerabilities and lived experience, underscoring that misalignment is not merely a matter of information deficits.

Overall, the results tell a clear and urgent story: extreme heat risk is rising faster than public awareness, and past experience alone is a poor guide to future danger. The RAP framework provides a practical, data-driven way to identify “danger zone” communities—places where heat risk is high but concern is low—and to target climate risk communication, public health outreach, and adaptation investments where they are most needed.

The full open-access article is available here at the journal Nature Communications. The interactive maps, more details on the tool, and responses to frequently asked questions can be found on our website here.

Citation

Marlon, J.R., Begotka, N., Preston, A., Steifel, M., Keenan, J.M., Inman, C., Malmin, N.P., Carman, J., Goddard, E., Kotcher, J. and Maibach, E., 2025. Experience-driven perceptions misalign with assessed heat risk in the United States. Nature Communications. https://doi.org/10.1038/s41467-025-67631-6

Funding Sources

This project was supported by the Schmidt Family Foundation, U.S. Energy Foundation, and the Heising-Simons Foundation.