Why we reviewed this paper
Most research on keeping cities cool comes from countries outside the Gulf region. Yet Gulf cities need it more than most. They also have something to offer that few other cities have: citywide irrigation networks that use treated sewage effluent (wastewater cleaned for reuse). We have found this has significant implications for cooling Gulf cities. These reviews explain academic research in plain English. They also point out where Gulf conditions would make a significant difference to the research, so readers don’t import ideas without taking the local context into account.
This is the first study to estimate how much trees cool the air in every large city in the world, Gulf cities included. Two of its findings bear directly on the Gulf. Trees cool more in dry climates. And the authors name water as the limit on planting more. Both are areas where irrigated Gulf cities differ from the cities behind most of the evidence.
The review
The short version
Cities are hotter than the countryside around them. Trees cancel just under half of that extra heat in large cities worldwide. But the cooling is small next to the warming climate change is expected to bring by 2050, and it goes mostly to richer, leafier places.
Why did the researchers do this?
On a summer evening, a city centre often feels warmer than the fields outside town. Concrete and asphalt soak up the sun and give the heat back out slowly. This matters because heat kills: an estimated 356,000 people died from heat worldwide in 2019. Trees cool streets with shade and by releasing water from their leaves. Until now, nobody had measured how much trees cool the air in every large city in the world, or compared that with future warming.
What did they do?
The team looked at all 8,919 large urban areas on Earth, home to about 3.6 billion people. They combined satellite maps of tree cover with estimates of summer air temperature, cut into squares about 1 km across. A computer model then estimated how hot each area would be with no trees, and with as many extra trees as could realistically be planted. They compared these results with climate projections for around 2050 under a “middle of the road” emissions path.
What did they find?
- Without trees, the gap between cities and their surroundings would be about twice what it is now.
- Trees cool city air by about 0.15°C on average. In some neighbourhoods the cooling reaches 2.7°C, and in many it is close to none.
- Climate change is expected to add about 1.5°C to summer days in cities by mid-century. Today’s trees offset about a tenth of that. Planting every tree that could plausibly fit would roughly double city tree cover and offset about a fifth.
- The benefits are unequal. In richer countries, about 4 in 10 city dwellers get noticeable cooling from trees. In poorer countries, fewer than 1 in 10 do.
- Suburbs have about twice the tree cover of dense city centres, even though more than half of city dwellers live in the dense parts.
- In dry, hot cities, each extra tree cools the air more than it would in a humid city.
What does this mean for you?
If you live in a crowded, hot city centre, especially in a poorer country, you are likely to get the least cooling from trees. The authors say planting should focus on these dense, low-income areas and on dry cities, as far as water allows, because that is where new trees do the most good. Trees help, but they cannot replace cutting greenhouse gas emissions to slow warming.
What changes in the Gulf
- Water is less of a limit. The authors recommend planting in dry cities as far as water resources allow. They suggest that irrigated trees in dry climates may release a lot of water from their leaves, which adds to the cooling. Gulf cities with irrigation networks fed by treated sewage effluent have a water source most dry cities lack. That makes the paper’s best case for planting, dry cities with water, a closer fit to the Gulf than to the arid cities it mostly draws on.
- The “realistic” planting limit may be too low for the Gulf. The authors set the most trees a neighbourhood could have by looking at the leafiest similar neighbourhoods that already exist. In dry climates, few cities are irrigated, so that ceiling is low. An irrigated Gulf city could go beyond it. The paper’s figure of trees offsetting about a fifth of future warming may understate what is possible here, but the paper did not test this.
- “Dry” here does not mean dry air. The paper classes climates by how little rain falls compared with how much water evaporates. Coastal Gulf cities are dry by that measure but very humid in summer. Trees cool by releasing water vapour, and in muggy coastal air the extra moisture may offset part of the cooling people feel. The paper’s test of how hot people actually feel covered Phoenix, Lisbon and Gothenburg. None of these is hot and humid, so its results should not be assumed for Gulf coastal cities.
- Wealth does not guarantee tree cooling. The paper finds richer countries get more cooling, but it groups countries by income. Gulf states are high-income and dry. The paper lists Saudi Arabia among the countries with the least cooling from trees. In the Gulf, climate and water matter more than income in deciding how much cooling trees provide.
What we still don’t know
The 1 km squares hide big differences from one street or garden to the next. The study measured air temperature only. A measure of how hot people actually feel, which includes humidity, sun and wind, was tested in just three cities. In those cities, it showed about three times more cooling from trees. The results are model estimates based on 2015–2019 data, not direct measurements in every city. The research was funded by The Nature Conservancy, Lyda Hill Philanthropies and the US Department of Energy, and the authors declare no competing interests.
Reference
McDonald, R. I., Chakraborty, T. C., Endreny, T. A., Parsons, L. A., Marsagishvili, M. & Esperon-Rodriguez, M. (2026). Trees halve urban heat island effect globally but unequal benefits only modestly mitigate climate-change warming. Nature Communications 17:3569. https://doi.org/10.1038/s41467-026-71825-x