Why Are Cities So Hot?
― The gap is bigger at night than by day
Same day, same weather — yet the city centre and the suburbs can differ by several degrees. And the gap is widest not at midday, but at night. Cities aren't hot because "sunlight hits harder." It's because they can't let go of the heat they've stored, once night falls.
Walking through a city on a summer night, hours after sunset, you can still feel heat rising from underfoot. Hold your hand near the asphalt — it's still warm.
Yet the same night, near a park or paddy field out in the suburbs, the air feels cool. Same city, wildly different feel depending on where you stand.
The temperature records back this up. The gap between suburb and city is bigger for the pre-dawn low than for the midday high.
Everywhere gets hot by day. A city's real problem is that it doesn't cool down.
Soil and plants shed heat by evaporating water. Asphalt and concrete have no water to evaporate. With no escape route, the heat simply piles up.
On a clear night, the ground radiates heat out into space. But surrounded by buildings, the visible patch of sky is small, and the heat it throws off bounces back off the walls.
On top of this, add the heat pumped out by air conditioners and cars. More heat comes in than goes out. That difference is the city's heat. Let's take it step by step.
The biggest factor: no water
On a clear day, the sun's energy reaching the ground has two possible fates: it warms the ground and air, or it evaporates water.
When water evaporates, it draws a huge amount of heat from its surroundings. It's exactly the same effect that cools you when sweat dries. On soil or grassland, a good chunk of the incoming energy goes into this evaporation. That's why the temperature doesn't climb as much.
Asphalt and concrete, though, have no water to evaporate. Even when it rains, the water runs off into the drains and is gone. With nowhere else to go, all that energy goes straight into raising the temperature.
On a clear summer day, asphalt surfaces are said to reach over 60°C. It gets too hot to touch not simply because it's dark-coloured, but because it has no mechanism to cool itself.
City ground can't.
Why heat can't escape at night
On a clear night, the ground radiates heat out toward space. This is radiative cooling — the same process behind frost on a winter morning.
But there's a condition: the sky has to be visible. Heat that's thrown off bounces back if it hits clouds (which is why cloudy nights don't cool much). The same happens if it hits a building.
Look up from a street hemmed in by buildings, and all you see is a narrow strip of sky. Most of the heat trying to escape gets absorbed by the surrounding walls and sent back. And since those walls have themselves warmed up during the day, they radiate heat right back at you.
On top of that, city buildings block the wind. The breeze that would carry the heat away can't get in.
The result: cities don't cool down even after night falls. That's why the urban heat island gap is larger at night.
An air conditioner doesn't destroy heat — it moves the heat from inside the room to outside. And since it uses electricity to do the moving, that electricity also ends up as heat released outdoors. The heat coming out of the outdoor unit is always more than the heat it removed from the room.
One unit barely matters. But when every air conditioner in the city runs at once, it adds up. Heat drives people to use air conditioning, and that waste heat makes the city even hotter — a real feedback loop.
To be clear: this isn't an argument for going without air conditioning. Heatstroke can be fatal. Using it properly indoors comes first — tackling the city-wide problem (more greenery, different pavement, smarter waste-heat management) is a separate matter.
How much hotter, exactly
Long-term temperature records make the urban rise stand out clearly.
Tokyo's annual average temperature has risen by roughly 3°C per century, it's said. Average that over sites with little urbanisation, though, and the rise is only about 1.5°C.
That gap is thought to come from urbanisation itself. In other words, city dwellers face an extra layer of heat on top of global warming as a whole.
The change is especially large in the overnight low. Nights that never drop below 25°C ("tropical nights") have clearly increased in major cities. Not being able to cool down while asleep raises the risk of heatstroke.
Effective measures follow from the mechanism
- Add greenery ― Trees and lawns evaporate water and cast shade. Because they "sweat" as well as shade, they cool more than a plain roof or awning
- Paint roofs and walls white ― Reflect sunlight and the heat never gets in to begin with. High-reflectivity coatings are said to sharply lower roof temperatures
- Use water-retaining pavement ― Pavement that soaks up rainwater and lets it evaporate slowly is already in use. It's a way of making the ground sweat
- Sprinkle water on the ground ― The most direct way to use evaporative cooling. Sprinkling it on sunlit ground at midday dries up fast and the effect is short-lived, so shaded spots in the morning or evening work better
- Keep wind corridors open ― Using rivers and wide roads as channels for cool air to flow in. Building layout plays a role here
Conversely, adding more buildings to create shade doesn't necessarily cool things down. You get shade, but heat also struggles to escape at night. The ideal shape differs between day and night — that's what makes this problem hard.
Something you can check outside
- On a clear afternoon, head outside (don't forget a hat and water)
- Hold your palm near sunlit asphalt (careful — don't touch, it's hot)
- Do the same over nearby sunlit soil or grass
- Try shaded asphalt and shaded soil too, and compare all four
- If you can, repeat at the same spots two hours after sunset
In the sun, asphalt and grassland feel startlingly different. But the night-time check is the most striking. The grass has cooled right down, while the asphalt is still warm — it's in the middle of releasing the heat it stored up during the day. Keep outdoor observation brief on hot days and watch how you feel. If you suspect heatstroke, move somewhere cool, take on water and salt, and if symptoms are severe, don't hesitate to call 119.
Summary
Cities are hot because three things stack up: ① the ground can't evaporate water, so incoming heat becomes temperature directly; ② at night, hemmed in by buildings, heat can't escape to the sky; and ③ waste heat from air conditioners and cars adds on top. It's especially ② that makes the gap larger at night.
Cities aren't hot because of the daytime.
They've lost their chance to cool down.
For a closer look at just how hot paved ground itself can get under the same sunshine, see The Air Is 32°C — So Why Does Asphalt Top 60°C?
Want to go deeper? ― Terms, numbers, and where this fits in the textbooksFrom middle-school science to open research questions, each labelled by level
- MSCovered in middle-school science
- HSCovered in high-school "Basic Physics" / "Basic Earth Science"
- HS+Covered in high-school "Physics" / "Earth Science," or treated as advanced/sidebar material in textbooks
- Univ.Not covered in high school — a university-level specialist topic (urban climatology)
- ResearchNot yet settled even at university level — something researchers are actively studying
MSTerms: words for describing urban heat
- Urban heat island: the phenomenon where a city's temperature runs higher than its surroundings. The name comes from how a city appears as an "island" when isotherms are drawn on a map.
- Latent heat of vaporization: the heat that water draws from its surroundings as it evaporates. It's why sprinkled water and sweat feel cooling.
- Radiative cooling: the ground losing heat by radiating it as infrared out into space. Strongest on clear nights.
- Tropical night: a night whose minimum temperature stays at or above 25°C. Clearly increasing in cities.
- Albedo: the fraction of sunlight a surface reflects. Higher for lighter colours, lower for darker ones.
HSDo the maths: how much sunlight does one litre of sprinkled water cancel out?
Does sprinkling water actually work, or is it just a comforting ritual? Put a number on "how much it helps," and you get an answer. All we need is how much heat water draws away as it evaporates.
Heat removed = latent heat × amount of water
| Heat removed | units of J (joules) |
| Latent heat | about 2260 J per gram of water |
| Amount of water | units of g |
When water evaporates, it carries heat away from its surroundings. This is a different mechanism from "cooling with cold water." Same as sweat cooling you down — it's not a drop in temperature that does the work, but the act of evaporation itself carrying heat away.
2260 J is an unusually large number among water's properties. It's over 500 times the heat needed to raise the same amount of water by 1°C.
| Water sprinkled per m² | 1 L = 1000 g |
| Heat removed | 2260 × 1000 = 2,260,000 J |
| Midsummer sunlight | about 1000 W/m² (1000 J per second) |
| How many seconds' worth | 2,260,000 ÷ 1000 = 2260 seconds |
| Converted to minutes | 2260 ÷ 60 ≈ 38 minutes |
One litre over 1 m² cancels out roughly 38 minutes of midsummer sunlight. Not just a comforting ritual. But it is only 38 minutes — it doesn't last all day.
That's why sprinkling water "in the evening, once the sun has lowered" makes sense. When sunlight is weaker, that same 38-minute effect stretches further. Sprinkle it at high noon and it just evaporates and is gone. The traditional practice matches the calculation.
Another big factor is how much sunlight the ground reflects back. Dark surfaces absorb it; pale surfaces bounce it away.
| Reflectance of ordinary asphalt | about 0.1 |
| Fraction absorbed | 1 − 0.1 = 0.9 |
| Heat absorbed | 1000 × 0.9 = 900 W/m² |
| Reflectance of light-coloured pavement | about 0.4 |
| Heat absorbed | 1000 × 0.6 = 600 W/m² |
| Difference | 900 − 600 = 300 W/m² |
Changing colour alone makes a 300 W difference per square metre. That's three-tenths of the incoming sunlight. And unlike sprinkling, you don't have to keep doing it.
Compare it with sprinkling water. Matching that 300 W/m² with water alone would need 2,260,000 ÷ 300 ≈ 7533 seconds — in other words, sprinkling a litre every two hours, continuously.
This shows the gap in scale between what an individual can do (sprinkling water) and what a city can do (pavement, greenery). Sprinkling works, but only briefly. Changing the ground itself does far more to shift the city's overall heat.
※ Sunlight levels and reflectance vary widely with weather and materials, and in practice wind and building shadows matter too. This calculation is meant to give a sense of scale.
HSThe numbers: city vs. suburb
| Asphalt reflectance | about 0.05–0.1 (absorbs over 90% of incoming light) |
| Concrete reflectance | about 0.2–0.3 |
| Grassland reflectance | about 0.2–0.25 (but also cooled by evaporation) |
| White-painted roof | about 0.6 or higher |
| Asphalt surface temperature on a clear summer day | can exceed 60°C |
The effect of sprinkling water can also be calculated. The heat water draws away as it evaporates is roughly 2,400 joules per gram.
| Heat removed | 2,400 × 1,000 = 2.4 million J (2.4 MJ) |
| Midsummer sunlight (per square metre) | roughly 1,000 W = 1,000 J per second |
| Time it can offset | 2,400,000 ÷ 1,000 = about 2,400 seconds (roughly 40 minutes) |
※ In practice this depends on wind and humidity. A rough estimate to give a sense of scale.
This calculation reveals sprinkling's nature: it works, but doesn't last. That's why it's said to be more efficient in shade, or in the morning and evening, rather than in full midday sun — slower evaporation stretches the cooling time out longer.
HS+Balancing the heat that comes in against what goes out
The energy budget at the ground surface can be written out as follows.
(incoming sunlight) − (reflected) − (radiated away as infrared) = (heats the air) + (evaporates water) + (stored in the ground)
In the suburbs, the "evaporates water" term on the right is large, so "heats the air" stays small. In the city, evaporation is close to zero, so the same energy is split between "heats the air" and "stored."
The ratio of "heats the air" to "evaporates water" is called the Bowen ratio. This ratio is high in cities and low in the suburbs. A single number captures the essence of why cities run hot.
And the "stored" portion is released at night. Asphalt and concrete store heat well and conduct it readily, so they warm deep down and keep radiating heat all night long.
Univ.How much sky is visible ― sky view factor
In urban climatology, the fraction of the sky visible as a solid angle from a given point is called the sky view factor, and it's a key indicator for explaining the strength of the urban heat island.
On open grassland the sky view factor is close to 1, and radiative cooling works fully. In a street hemmed in by buildings (a "street canyon"), it can drop below 0.3, and much of the infrared thrown off gets absorbed by the walls and sent back. A clear relationship has been observed between night-time heat island intensity and sky view factor.
The urban atmosphere behaves distinctively within a layer a few hundred metres deep near the surface (the urban boundary layer). It mixes vigorously by day, and at night it rarely forms the stable layer that develops over the suburbs. In the suburbs, cool air pools near the ground at night; in the city this rarely happens.
The temperature gap between city and suburb also generates its own local wind. Rising air over the warm city draws air in from the suburbs, a circulation that also affects how pollutants spread.
ResearchWhat's still unresolved
- Cleanly separating the urbanisation share from the global-warming share isn't simple. The main text cites "about 3°C for Tokyo, about 1.5°C at sites with little urbanisation" — but even choosing those "little urbanisation" sites involves judgment calls. The surroundings of any weather station also change over decades, so long-term records completely free of urban influence are hard to find. Debate over how to separate the two continues.
- How cities affect rainfall is also unsettled. There are suggestions that a city's rising air, waste heat, and particulates intensify localised heavy rain. Some reports find increased rainfall downwind of cities, but conditions differ city to city, making it hard to distinguish from natural variability.
- Quantitatively predicting the effect of greening is also still developing. Planting trees cools things down, that much is certain — but where, how much, and in what arrangement determines the actual temperature drop, and that's heavily shaped by surrounding buildings and airflow. Research into finding the arrangement that gives the biggest effect for a limited budget is ongoing, but there's no standard answer yet.
- The health impact of heat isn't determined by temperature alone. Humidity, wind, sunlight, and people's behaviour and living conditions all play a part. Predicting "at what temperature, who is at risk" is still not precise enough. The effect on the body of overnight temperatures that never drop is an area especially in need of more research.
Where this fits in the textbooks (by level)
| Level | Subject / unit | Where in this article |
|---|---|---|
| MS | Science: states of matter and heat / weather / solar energy | Latent heat, radiative cooling, observing ground temperature |
| HS | Basic Physics: heat and specific heat / Basic Earth Science: atmosphere and energy balance | Reflectance, the sprinkling calculation, heat storage |
| HS+ | Physics: thermal radiation / Earth Science: radiation budget | The energy-budget equation, Bowen ratio |
| Univ. | Urban climatology / building environmental engineering | Sky view factor, urban boundary layer, local circulation |
| Research | Climatology / urban planning (unresolved) | Separating contributions, cities and rainfall, optimal greening layout |
| ― | Public health | Tropical nights and heatstroke, effects of not cooling down at night |
- Japan Meteorological Agency, "Heat Island Monitoring Report" (気象庁「ヒートアイランド監視報告」) and its analysis of urbanisation effects.
- Oke, T. R., Boundary Layer Climates and Urban Climates (standard textbooks of urban climatology; sky view factor and night-time heat islands).
- Ministry of the Environment (Japan) heat island countermeasure guidelines (環境省「ヒートアイランド対策ガイドライン」: greening, high-reflectivity coatings, water-retaining pavement).
- Oke, T. R., City size and the urban heat island, Atmospheric Environment 7, 769–779, 1973.
- Ministry of Health, Labour and Welfare and Ministry of the Environment (Japan) information on heatstroke prevention.
※ Figures for temperature rise, reflectance and the like vary by location and conditions. This article shows commonly cited approximate values.
※ This article is a general-audience science explainer. For guidance on coping with heat, follow information from the Ministry of the Environment, the Ministry of Health, Labour and Welfare, and your local government. Keep outdoor observation brief, and if you notice anything wrong with how you feel, move to a cool place and contact a medical facility or call 119 if needed. The figures given here are approximate, meant to help you understand the underlying mechanism.