How to prioritize cool pavement pilot streets on a limited budget

Most cities don't get to coat every arterial at once. A cool pavement pilot usually runs on a few hundred thousand dollars and a handful of lane-miles, which means the hardest part of the project isn't the coating spec. It's deciding which blocks get it first, and being able to defend that choice to a budget committee that wants to know why Block A beat Block B.

Start with what's hot, not what's assumed to be hot

Planners often default to the obvious candidates: downtown arterials, industrial corridors, the strip mall parking lots everyone complains about in July. Some of those are genuinely the worst offenders. Others just look hot because they're visible and paved in dark asphalt that's a few years old.

The problem with picking by instinct is that two streets can look identical from a windshield survey and still have very different surface temperatures, because the age of the asphalt, the mix, and the amount of shade overhead all change how much heat a block holds by 3pm. A block with a thin, newer seal coat can run cooler than an older, oxidized stretch two streets over, even though both read as "standard asphalt" on the GIS layer.

This is where pulling in a thermal read of the street network pays for itself before a single dollar of coating gets spent. Looking at surface temperature next to the material underneath it, rather than guessing from pavement age records or a drive-by, turns a shortlist of 40 candidate blocks into the dozen running hottest for their material type.

Match the hottest blocks to streets you're already touching

Capital budgets stretch further when the cool pavement treatment rides on a resurfacing project that's already funded. If public works has a mill-and-overlay schedule for the next two construction seasons, cross-reference that list against the blocks with the worst surface temperatures. A street that's already getting new asphalt is a cheap add for a cool coating; a street that would need its own standalone project is a much harder budget ask.

This overlap also solves the "why this block and not that one" question before it comes up in council. If the answer is "it was scheduled for resurfacing anyway and it happens to run hotter than comparable streets nearby," that's a defensible, boring, correct answer. Boring is good when you're asking for capital money.

A few other filters worth running against the shortlist before locking in pilot sites:

• Pedestrian exposure at peak heat hours, not just traffic volume. A collector street with heavy foot traffic near a transit stop matters more than a wider arterial that's mostly cars passing through.
• Canopy cover on the block. Streets with decent tree cover may already be getting some relief, and the coating's temperature drop is more noticeable on streets with little shade to begin with.
• Distance between candidate blocks, if you want before-and-after comparisons that aren't contaminated by a neighboring treated or untreated street skewing the read.

Keep the pilot narrow enough to measure

A pilot that spreads across eight unrelated neighborhoods is hard to evaluate at the end of the season, because you're comparing blocks with different baseline materials, different ages, different canopy, and different traffic patterns all at once. A tighter pilot, four to six blocks that share enough characteristics to compare cleanly, gives a city a result it can point to when asking for the next round of funding: this material, this surface temperature drop, this cost per block.

For that kind of before-and-after comparison, having a baseline map that ties surface temperature to pavement material at the block level, rather than a single citywide heat average, matters as much after the pilot as before it. Urban Heat Island builds that map from thermal and multispectral satellite passes each season, so the same shortlist process that picked the pilot streets can also show whether the coating moved the needle on the blocks that got it.

If the next capital cycle needs a shortlist that holds up under questioning, start with a block-level read of where the heat actually is.