Have a question?
As cities grow around weather stations, does the hot pavement skew our measurements of global warming?
No—climate scientists are careful to remove the signal of urban warming from the global temperature record. Even in the raw, unadjusted data, urban heat islands explain a very small fraction of the global climate change we have observed.
June 10, 2026
Measuring our planet’s temperature is a vast effort, relying on a worldwide network of weather stations, satellites, ships, and buoys. With all these sources taking regular measurements, scientists can piece together a complete picture of Earth’s surface temperature. This work has provided the most direct evidence of climate change, documenting that, over the past century, the planet has been steadily heating up.
On its own, however, this temperature record can’t tell us why the Earth is warming. To be sure that the buildup of heat-trapping greenhouse gases in our atmosphere is the dominant cause of climate change, scientists need to account for other possible explanations. For instance, as cities expand around the world, weather stations that were built amid fields and forests may end up in a maze of highways, rooftops, and parking lots. Because the resulting loss of plant life and growth of dark surfaces can make cities significantly warmer than the surrounding countryside, it’s reasonable to wonder whether their ongoing sprawl might explain the rise we see in the temperature record.
These “urban heat islands” are real and important, says Adam Schlosser, a climate modeler and deputy director of the MIT Center for Sustainability Science and Strategy—especially for city dwellers who are more at risk from extreme heat as the planet warms. But climate scientists have put a lot of thought and effort into pinpointing how much cities affect our measurements of rising temperatures. “There’s a very clear warming signal that can’t be explained by the growth of urban environments,” he says.
This is easy to see just by looking at a map of Earth’s temperature change. Global warming, Schlosser points out, is not concentrated in and around cities. Weather stations in rural environments have also recorded rising temperatures for over a century. So have temperature readings from the ocean surface, which is obviously not being paved over. In fact, some of the places that are warming fastest, like the Arctic, are also the furthest from urban sprawl.
Still, says Schlosser, while it’s clear at a glance that cities can’t explain the whole warming trend, climate scientists do need to find and measure urban heat islands. Otherwise, we might overstate the warming caused by global factors—most notably, by humanity’s greenhouse gas emissions.
“When you think about where weather stations have been taking records over the course of 100 or 150 years, many of them were built in environments that weren’t urbanized to begin with,” Schlosser says. “The urban heat island effect is an important signal that we need to extract, so that we can definitively and accurately assess other warming effects globally.”
The different organizations that measure the Earth’s temperature—among them, NASA, the UK Met Office, and the European Union’s Copernicus program—have different methods for dealing with urban heat islands, but they all share the same basic ideas. First, scientists need to identify the weather stations affected by urban heat islands. NASA, for example, sorts out urban from rural weather stations using nighttime brightness, measured by satellites.1
Once scientists have spotted these stations, they have to clean up their temperature data, to untangle the local effects of urban warming from the larger trend of global climate change. The simplest option is to drop the urban stations from the record entirely, and some analyses do just that.2 It turns out this works quite well, because even in the most developed regions of the world, there is plenty of rural land to fill the map. (Consider that urban areas make up just 3% of the United States.3)
In the NASA method, however, scientists keep the urban stations, but adjust their readings. To do this, they compare each urban station with every rural station within 1,000 kilometers, and rebalance the urban stations’ long-term warming trends to match those of their rural neighbors.1,4 Just like dropping the urban stations, this process reveals that climate change has warmed the Earth slightly less than the raw data would suggest. But adjusting rather than excluding the urban data is helpful in other ways. The NASA record can keep the day-to-day ups and downs of the urban stations, and can also more easily extend backward into the 1800s, as many of the oldest weather stations are in cities and suburbs.
Notably, these adjustments make only a minor difference in calculations of the Earth’s temperature rise. NASA has found that, whether they use their adjustment method, use the raw data from all weather stations, or use only the most pristine rural stations that are pitch-black at night, their estimate of how much the world has warmed since 1900 changes by only around 0.01°C (0.02°F).1 In a world that has already warmed by well over 1° C, that’s a difference of less than 1%. Despite their swift growth, cities and suburbs cover just a tiny fraction of the planet, so while their local heating effects can be strong, they are dwarfed at the global scale by our greenhouse gas emissions.
“Fortunately, we haven't paved enough of our planet for it to make an impact globally,” says Schlosser.
Nonetheless, it’s good science to do this cleanup. Even with rich temperature data at hand, scientists have to work hard to make sure they’re reading it properly—adjusting not only for cities, but also for the time of day temperatures are taken, how spread out readings are around the world, and other variables that could skew our measurements. Once these features are accounted for, we can see the strongest signal of what’s happened to our climate since the rise of modern industry: a worldwide pattern of warming that is faster and more widespread than anything human civilization has previously encountered.
We are grateful to Nathan Lenssen, Assistant Teaching Professor in Applied Mathematics and Statistics at Colorado School of Mines and Scientist at the NSF National Center of Atmospheric Research, for additional assistance with this article.
Submit your own question to Ask MIT Climate
Get the latest from Ask MIT Climate monthly in your inbox
1 Hansen, James, et al. "Global surface temperature change." Reviews of Geophysics 48 (2010). https://doi.org/10.1029/2010RG000345.
2 Parker, David. "Urban heat island effects on estimates of observed climate change." Wiley Interdisciplinary Reviews Climate Change 1 (2010). https://doi.org/10.1002/wcc.21.
3 U.S. Department of Agriculture Economic Research Service: Major land uses. Updated March 13, 2026.
4 Hansen, James, et al. "GISS analysis of surface temperature change." Journal of Geophysical Research Atmospheres 104 (1999). https://doi.org/10.1029/1999JD900835.
Want to learn more?
Listen to this episode of Ask MIT Climate on the scientific challenge of measuring global temperatures.
Keep exploring
Check out these related Explainers, written by scientists and experts from MIT and beyond.
