Urban heat islands
A Socratic walk-through of urban heat islands — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why is a city several degrees warmer than the countryside surrounding it?
Drive out of a city on a summer evening with the car thermometer visible and you can watch the number fall — a degree, two, sometimes more, over a few miles of darker fields.
The obvious explanation is that cities make heat: engines, air conditioners, furnaces, bodies. That is real, and we will come back to it. But it cannot be the main story, because the difference is largest at the hour when the city is quietest, and smallest in the middle of a working afternoon. Something is wrong with the intuition. What, exactly?
Reasoning it through
REASONING #Start not with heat production but with the energy budget of a patch of ground. Sunlight arrives; some bounces straight back, and that fraction is the albedo. What is absorbed has to go somewhere: it warms the surface, conducts into the material, passes to the air as sensible heat, evaporates water, or radiates back to the sky as infrared. Every difference between a city and a field is a change to one of those terms.
Albedo first. Asphalt, dark roofs and weathered concrete reflect less than grass or bare soil, so more sunlight is absorbed to begin with. Now the water term. A field is full of plants moving water from soil to air, and evaporation is expensive — turning liquid water into vapour takes a great deal of energy, and that energy goes into the vapour rather than into raising temperature. A city drains its rain into pipes within minutes and has little vegetation left to transpire, so the same absorbed sunlight has fewer places to go and more of it ends up as sensible heat.
Third, the materials. Brick, stone, concrete and asphalt store heat readily and conduct it inward, so a city spends the day charging a very large thermal battery — with far more surface to charge, not just ground but walls, several storeys of them, facing the sun.
Now ask what happens after sunset, because that is where the puzzle sits. With the sun gone, every surface cools by radiating infrared to a very cold sky. A flat field has the whole hemisphere above it. A street between two buildings does not: most of the sky is blocked, and what a wall radiates upward is largely intercepted by the wall opposite, which radiates it straight back. The proportion of open sky a point can see is the sky view factor, and in a narrow street it can be well under half. The city has not stopped cooling; it is cooling slowly, while the countryside cools fast. Add the stored daytime heat now bleeding out of the masonry, and buildings slowing the wind that would mix warm air away, and the gap opens through the evening rather than closing.
That is why the effect peaks at night, a few hours after sunset, and why it is largest on calm, cloudless nights — cloud blocks the countryside's escape too, and wind mixes the difference away. Only now does waste heat belong in the account: vehicles, buildings, industry and air conditioning add a genuine extra term, large enough to matter in dense cores and cold-city winters, but a contributor rather than the engine.
One more distinction, the one most often muddled. Measure the surface with a thermal camera at midday and a car park may be far hotter than a lawn; measure the air at head height and the same difference may be nearly nothing. Those are two different heat islands — the surface one a daytime phenomenon, the air-temperature one a night-time phenomenon — and photographs of the first routinely illustrate the second.
The analogy
THE ANALOGY #Consider two stone jars taken from the same oven, one left on an open table and the other set inside a deep, narrow box open only at the top. Both are equally hot, and both lose heat only by radiating it. The jar in the box radiates into walls nearly as warm as itself, and gets most of it back. Come back at midnight: the jar on the table is cool, the jar in the box is still warm — and nothing was ever added to it.
the jars start equally hot, whereas a city genuinely absorbs more heat by day than a field does, so it begins the evening ahead as well as cooling more slowly. And the box is fixed geometry; a real street also has wind, its own waste heat, and moisture moving through it.
Clarifying the model
THE MODEL #The refinement that makes the pieces cohere is that "the city is hotter" is really two claims: it takes on more heat, and it lets go of it more slowly. Albedo and evapotranspiration govern the first; sky view factor, thermal mass and reduced wind govern the second; waste heat adds to both. The night-time peak follows because the loading terms stop at sunset while the release terms keep running.
That also explains what helps. Cool roofs and lighter pavements attack the albedo term, so their benefit is mostly a daytime one. Trees attack two terms at once — shade before absorption, transpiration after — which is why they generally outperform surface treatments, though a canopy also blocks some night sky and can slightly slow cooling.
Two honest limits. The magnitude is very variable, depending on city size and density, climate, season, and above all the weather that night, so a single figure for "how much warmer a city is" means little. And attributing the effect to any one mechanism at a given site is hard, because they are entangled: dense street canyons come with low albedo, little vegetation and concentrated waste heat all at once.
A picture of it
THE PICTURE #How to readThe curve is the city's air temperature minus the countryside's, hour by hour — so it is a difference, not a temperature, and zero would mean the two are identical. Read the shape rather than the numbers, which are schematic: the actual size varies enormously with city and weather. Notice the trough in the early afternoon, when both places are hot and the gap nearly closes, and the steep climb after sunset around hour 18. The peak sits in the late evening, hours after the traffic and the sunshine have gone — which is the whole argument. The gap is not made by the city producing heat; it is made by the city failing to lose it.
What became clearer
WHAT CLEARED #A city is warm at night because of what it cannot do, not what it does. Dark, dry surfaces absorb more of the day's sunlight and store it in walls and roads; with no vegetation spending energy on evaporation, more of that heat reaches the air; and after dark the street canyon blocks the sky it needs to radiate into. Waste heat is a real addition, but the effect would exist in a city with the engines switched off. The clearest sign that the mechanism is the cooling and not the heating is the timing: the difference is smallest in mid-afternoon and largest hours after everyone has gone home.
Where to go next
ONWARD #- Why the same city can show a heat island in winter for entirely different reasons, dominated by waste heat.
- How urban heat and night-time humidity combine to affect mortality more than either does alone.
Key terms
TERMS #| Term | What it means |
|---|---|
| Albedo | the fraction of incoming sunlight a surface reflects rather than absorbs. |
| Evapotranspiration | the transfer of water to the air from soil evaporation and plant transpiration, which consumes energy that would otherwise heat the air. |
| Sensible heat | energy transfer that raises temperature, as opposed to latent heat, which changes water's phase instead. |
| Sky view factor | the fraction of sky visible from a point; low in a narrow street, which slows night-time radiative cooling. |
Every term the collection defines is gathered in the glossary.