Surface urban heat island vs canopy urban heat island: what's the difference

Two planners can look at the same city and describe completely different heat problems, and both can be right. One is talking about surface urban heat island. The other is talking about canopy urban heat island. They get used interchangeably in grant applications and council presentations, but they measure different things, at different heights, with different instruments, and a cooling plan built on the wrong one will miss the blocks that need it most.
Surface UHI: what the pavement is doing
Surface UHI is the temperature of the built surface itself: asphalt, rooftops, parking lots, bare soil, tree canopy from above. It's captured by satellite thermal bands reading radiant heat off whatever is facing the sky at the moment of overpass, usually midday or early afternoon when surfaces are at their hottest.
A big box store roof can read 150°F+ in July. A shaded residential block two streets over can read 40 degrees cooler on the same surface map, even though the air above both blocks feels roughly similar to a person walking through it. Surface UHI matters for where heat gets absorbed during the day and re-radiated into the evening. It's a skin-temperature reading taken from above, separate from what a person standing on that block feels at head height.
Canopy UHI: what people breathe
Canopy UHI refers to air temperature within the urban canopy layer, roughly street level up to rooftop height, the zone people live and move through. It's measured by weather stations, mobile sensor traverses, or networks of fixed air temperature loggers, not by satellite. Canopy UHI tends to peak later in the day than surface UHI, often after sunset, as heat stored in pavement and masonry during the day releases slowly into the air overnight. That's why a neighborhood with hot surfaces in the afternoon often stays hot at 9pm when a park two blocks away has already cooled off.
This is the number that correlates with heat-related illness. Nobody collapses from touching a hot parking lot from three stories up. People collapse from air that doesn't cool down enough overnight to let the body recover before the next day's heat starts stacking on top of it. Canopy UHI is the one that actually tracks that risk, and it requires ground sensors most cities don't have at block density, usually clustered near airports or a handful of research sites instead of spread evenly across every census block group.
Why the mismatch matters for block-level targeting
Surface UHI data is comparatively cheap to get at fine spatial resolution, because satellites already fly over every block on a repeating schedule. Canopy UHI is the harder number to collect at scale. So planners end up doing one of two things. Either they site interventions using sparse air temperature data that can't resolve block-to-block variation, or they use surface temperature as a proxy for heat exposure and risk getting the geography wrong, because a block with a hot roof and no residents is not the same priority as a block with a merely warm roof, no canopy, old building stock, and a lot of people over 65 without air conditioning.
The two measurements don't move in lockstep. A block can run hot on surface temperature and moderate on canopy temperature if it has strong ventilation. Another block can run moderate on surface temperature but brutal on canopy temperature if it's surrounded by heat-retaining masonry with no airflow at night. Treating surface UHI as a stand-in for lived heat exposure, without checking what's on the ground at that location, is how cooling-center dollars end up at the wrong blocks.
A planner needs to read surface temperature alongside building type, tree canopy cover, and who lives on each block, so the surface signal gets interpreted in context instead of taken as a direct read of resident risk. Heat Exposure Mapping layers seasonal surface temperature with building footprints and demographic density so a planner can see which specific blocks combine hot surfaces with vulnerable residents and thin tree cover, the combination that predicts heat strain on the ground, beyond hot pavement alone.
If you're deciding where the next round of cooling centers or shade trees goes, start by seeing which blocks carry that combination, not just which ones look hottest from above.