What is a heat vulnerability index? A block-level primer
A heat vulnerability index (HVI) is a composite score that tells a city which neighborhoods carry the highest combined risk from extreme heat. It's a different thing from a thermal map of which blocks simply run hottest. The index blends land surface temperature with demographic factors like age and income, and with built-environment variables like tree canopy and impervious surface cover. The output is a ranked or classed score, usually by census tract, that public health departments and planners use to decide where to put cooling centers, shade trees, and heat-emergency outreach.
Most HVIs trace back to a handful of CDC and NOAA-funded pilot studies from the 2010s, and dozens of cities have built their own version since. The method is public and well documented. The inputs, and the geography a city chooses to score, matter as much as the formula itself.
What goes into the score
A typical HVI pulls together:
- Land surface temperature, usually from satellite thermal bands, averaged over a summer period or a specific heat event
- Demographic density of vulnerable groups: people over 65, people under 5, people living alone, households without air conditioning or without a vehicle
- Housing and building stock, including older multifamily buildings with poor insulation, mobile homes, and buildings without central air
- Tree canopy and impervious surface cover, since asphalt and rooftops hold heat long after the sun sets and canopy is one of the few interventions that actually lowers it
Each factor gets normalized and weighted, then summed into a single score per geographic unit. Some cities rank by percentile. Some bucket into five classes from low to extreme. Either way, the point is the same: turn five or six layers of data into one number a budget committee can act on.
Why the geography you score matters
Census tracts are the default unit because the demographic data already comes packaged that way from the Census Bureau's American Community Survey. That's convenient. It isn't always accurate. A tract can run three or four degrees warmer on one side than the other, split by a highway, a lot full of surface parking, or a stretch of un-shaded apartment blocks sitting right against it. Average the whole tract and you can end up siting a cooling center or a tree-planting crew a half mile from the actual hot spot.
That's the gap between a tract-level heat vulnerability index and one scored at finer resolution. The scoring keeps the same inputs, temperature, demographics, building type, but assigns them street by street instead of averaging across the whole tract. A planner can see that one specific stretch is driving the tract's whole score, while the stretch next to it, with mature street trees and lower density, sits well below it. Heat Exposure Mapping overlays surface temperature, building type, and demographic density by block across a city, so the ranking points at the blocks that actually run hottest and house the most heat-vulnerable residents instead of an averaged tract that blurs the two together.
Using the score to target cooling investment
An HVI on its own doesn't plant a tree or open a cooling center. It's a prioritization tool, and prioritization only works if the units being ranked are small enough to match the budget line. A city with funding for twenty new shade trees doesn't need to know which of forty census tracts counts as high vulnerability. It needs to know which forty blocks, inside whichever tracts, would benefit most from canopy right now.
The same logic applies to cooling-center siting, to outreach ahead of a heat watch, to where a mobile misting unit gets parked on the worst day of August. The index is only as useful as the resolution it's built at, and the resolution should match the size of the intervention being planned.
If your city already has a vulnerability index built at the tract level, and the next question is which blocks inside those tracts deserve the next round of cooling investment, that's the question worth bringing to Heat Exposure Mapping.