About HeatLens
HeatLens translates satellite imagery into a single readable score — so planners, advocates, and residents can see exactly which communities bear the heaviest heat burden and where cooling investments will have the greatest impact.
HeatLens is an open urban heat risk tool covering the United States. It reads three satellite-derived datasets, combines them into a weighted score of 0–100, and maps the result by county, city, and census tract — so that anyone, whether resident, journalist, planner, or policymaker, can see at a glance which communities carry the heaviest heat burden and what drives it.
Urban heat islands form where concrete, asphalt, and buildings trap radiant energy that vegetation and soil would otherwise absorb or release through evaporation. Inner-city neighborhoods can run 7–10 °F hotter than nearby suburbs on a summer evening. That gap is rarely random: low-canopy, high-density districts were disproportionately shaped by decades of disinvestment. Extreme heat is now among the deadliest weather hazards worldwide.
Each area receives a score from 0 (severe heat risk) to 100 (coolest). Three satellite-derived factors are normalized, weighted, and summed. Higher contributions from temperature and impervious surface push the score down; more tree canopy pushes it up.
Score = 100 − (0.40 × T + 0.35 × (100 − C) + 0.25 × I)
T = surface temperature percentile within the surrounding region (0 = coolest, 100 = hottest) · C = tree canopy cover % · I = built-up surface %
At a score of 12, Chicago is the lowest-scoring — and highest-risk — city HeatLens has measured in the United States. The breakdown shows why: tree canopy sits at just 11.69%, while impervious surface covers 66.30%. That combination of minimal shade and heavy pavement is exactly the pattern HeatLens is built to surface.
HeatLens grew out of a gap in the research: major cities get studied, mapped, and funded, while the mid-sized and post-industrial towns where millions of people live are routinely overlooked. The Lansing case study behind HeatLens showed that the same heat dynamics shaping big metros are at work in these smaller cities too — they've simply never been measured. The vision is to close that gap: give any community, whatever its size or budget, the neighborhood-level heat intelligence that until now only major cities could access. Every place added is one more community that can see its risk and act on it.
It's easy to assume the hottest neighborhoods are simply the poorest ones — but the Lansing study behind HeatLens found heat didn't track income that way at all. The coolest neighborhood was the lowest-income one, while older, denser areas with less tree canopy ran hotter regardless of wealth. What drives exposure is the physical design of a place: its age, how much pavement it has, and how many trees were left standing. That's why HeatLens scores the built environment directly — the people most exposed to heat are whoever lives where canopy is thin and density is high, and that isn't always who you'd expect.
Cities, housing advocates, climate justice organizations, and researchers are welcome to collaborate for deeper analysis, custom area exports, or embedding HeatLens in public-facing planning tools. We're looking to collaborate with municipal equity offices, community development organizations, and university research groups.
Each area is scored 0–100, where 100 is coolest (lowest risk) and 0 is hottest (severe risk). Three satellite-derived factors combine to produce the score:
Scores are calculated for U.S. counties, states, cities, and neighborhoods (via Census tracts) using publicly available federal data. Scores are always displayed as numbers alongside color to remain legible for users with color vision deficiency.
HeatLens is a screening tool for identifying priority areas, not a substitute for site-specific, on-the-ground measurement.
We'd love to hear from you — whether you're a city or community organization exploring how HeatLens could support your work, a researcher interested in the data, or just have a question.