Urban Heat Islands and Climate Change

A recent story on NPR’s Morning Edition piqued my interest in the urban heat island (UHI) phenomenon, which has been recognized since the late 1980s. In the radio piece, the rather startling claim was made that cities are heating up at rates twice as fast as those measured for non-urbanized regions of the planet. The US city of Atlanta, GA is the focus of the NPR story, and Georgia Tech’s Urban Climate Lab director, Brian Stone Jr., is interviewed throughout, so I decided to take a look at the UCL website. The data on urban temperature trends are limited to fifty of the most populous US cities, but nevertheless support the need to develop mitigation strategies to counteract the impact of climate change on cities. For those who are interested in the comparison of warming trends in US cities vs. nearby rural areas, the climate trend map for the years 1961 through 2010 is here. Phoenix, Atlanta, Denver, and Greensboro are among the cities with the most alarming warming patterns.

Screen Shot 2012-09-12 at 7.58.01 AM

Atlanta skyline; photo by Chuck Koehler via Wikimedia Commons, under Creative Commons Attribution 2.0 Generic license

Many of us at OT live in large cities, whether in urban or suburban districts, and I was curious about how widespread the UHI phenomena of accelerated warming trends are worldwide. Turns out (not surprisingly) that there’s a huge amount of peer-reviewed literature available, and a variety of consequences for human health, as well as planned and implemented mitigation strategies, for cities across the globe. Wilby (2008) used data from studies by the London Climate Change Partnership to examine existing and projected changes in UHI intensity and ozone pollution in the city. UHI intensity is typically represented by a temperature gradient between an urban station (St. James’s Park in this case) and a rural reference station (Wisley, Surrey – 30 km from the city center), and the average nocturnal UHI for London is +2.2C in August. The factors that contribute to the UHI effect include building materials that retain more solar energy than do surfaces covered with vegetation, lower wind speeds, reduced evapotranspiration (due to a combination of fewer plants and relatively impermeable surfaces), and a concentration of anthropogenic heat flux (air conditioning, transportation, cooking). From long-term records, it’s clear that spring and summer nocturnal UHI intensities in London have increased since the 1960s. Using several different models, Wilby (2008) predicts that, by the 2050s, nightime UHI intensities for London will increase by another 0.5C for the months of May through October, and that there will be more frequent intense UHI episodes on a background of more persistent and extreme European heatwaves. This would mean, for example, that Londoners will experience an average of seven intense UHI episodes each August, compared with the current average of five.

What about the UHI effect in the large cities of Asia, Africa, and South America? Peng and colleagues (2012) analyzed seasonal and diurnal variations in UHI intensity for 419 global big cities, with populations over one million. Rather than relying on air temperature differences measured between urban and rural stations, the researchers used a land surface temperature (LST) data set from satellite remote sensing (MODerate-resolution Imaging Spectroradiometer = MODIS-Aqua) to compare urban and suburban areas, and thus calculate a Surface Urban Heat Island Intensity (SUHII). SUHIIs were computed for nighttime and daytime, and for winter and summer periods. Other parameters, including vegetation, climate, density, and albedo, were defined for urbann and suburban areas.

With the exception of a few cities surrounded by desert (e.g. Jeddah in Saudi Arabia and Mosul in Iraq), annual mean daytime SUHII is positive. Medellín, Colombia has the highest daytime SUHII (7.0C), followed by Tokyo and Nagoya in Japan, São Paulo in Brazil, and Bogatá in Colombia – all with daytime SUHIIs over 5C. Nighttime SUHIIs for most cities are between 0 and 2C, with Mexico City holding the record at 3.4C. Average daytime SUHII over cities in developed countries is higher than that over developing countries, whereas for nighttime SUHII there is no significant difference. The two major sources of energy that contribute to SUHII are downward net solar radiation and anthropogenic heat flux. During the day, evapotranspiration from urban vegetation has a cooling effect and reduces SUHII, whereas albedo and the thermal properties of urban surfaces influence nighttime SUHII. Interestingly, human metabolic heating accounts for only a very small fraction of anthropogenic heat flux – so don’t worry about going for a long run or playing tennis after work.

Obviously, the combination of global warming, an increase in extreme climate events, and accelerated urbanization requires that city planners consider measures to reduce SUHIIs. I hope to discuss a few of these strategies in subsequent posts.


Peng S, Piao S, Ciais P et al. (2012) Surface urban heat island across 419 global big cities. Environmental Science and Technology 46, 696-703.

Wilby RL (2008) Constructing climate change scenarios of urban heat island intensity and air quality. Environment and Planning B: Planning and Design 35, 902-919.

This entry was posted in climate change, urban landscape and tagged , , . Bookmark the permalink.

5 Responses to Urban Heat Islands and Climate Change

  1. My pet bugbear round where we live (urban suburb, mostly terraced housing with greenery, postage-stamp front and small rear gardens) is people paving the front gardens to stand their SUVs on, and the rear gardens too to create ostentatious patios to stand the barbecue, patio heater and loungers on. I estimate that in our row of a dozen houses only three have any grass left in the garden, and more than half barely even have a flower bed amid the paving.

    Apart from the (over) heating, it also directly increases the flooding risk – more of a problem than heating in the English North-West! – since the water has nowhere to run away too.

    • KristiV says:

      In Texas, our solution to the SUV parking problem is to design neighborhoods with large driveways and wide streets, and to sprawl further and further into the countryside. We are quite fond of large front and back lawns, which might counteract the paving effect somewhat, if we didn’t insist on dousing the grass with pesticides and fertilizers. I’ve considered replacing the front lawn with a xeriscaped rock garden, but the way such things are typically done around here, there would be too few plants and too much rock for my liking. I might go half and half, with part of the yard covered with more drought-tolerant grass and a few trees (currently there are two small yaupon hollies), and part as rock garden.

      The first place I lived in London (Tooting) was terraced housing, with no grass (or plants of any kind) in the front or back. We moved to a semi-detached in Clapham/Battersea, where the front yard was gravel, and the back was just a glorified patio with a few neglected plants around the edges. One of my flatmates was an architect, and he frequently bought cut flowers for the interior, but didn’t care at all about outdoor plants.

  2. I’ve heard this story often, about cities being hotter places than in rural areas. But my understanding is (from not having dug into the literature) that any climate effects this happens is local to the city/suburbs.

    The other story I’ve heard is that people living in cities have lower carbon footprints than people living in more rural areas. This is usually attributed to things like greater access to public transportation, and scaling efficiencies (for instance, it is more efficient to heat an apartment building of 10 units than to heat 10 stand alone houses of the same square footage).

    What I don’t know, and would love if you wanted to address, is how these two effects interacted with each other in terms of global climate change. On one level, having pockets of hot may be bad for total world temperature, but given that no one is thinking of some drastic population reducing measure, is it worse than any other possible population distribution scheme?

    • KristiV says:

      My understanding from the few papers I’ve read is that the heating effects are very local within cities, and the satellite remote sensing images bear this out. One of the studies that I hope to discuss in a future post addresses heat-related mortality in “micro-urban” heat islands within the city of Montreal.

      Two books on the green characteristics of cities might interest you (and I hope to review them here at some point – still in the midst of one of them): Edward Glaeser’s Triumph of the City: How Our Greatest Invention Makes Us Richer, Smarter, Greener, Healthier, and Happier and David Owen’s Green Metropolis: Why Living Smaller, Living Closer, and Driving Less Are the Keys to Sustainability. You can probably tell from the subtitles which side of the argument the authors favor. I’d love for others to comment on experiences and opinions based on living in cities across the globe … my only direct experiences outside the US are limited to London.

      • I have experience with one large city abroad, in the developing world, of about 4 million people. There are definite heat and pollution effects of living in the city. I have friends who need inhalers in the city, and can practically leave them at home when they are traveling elsewhere. During the summers, one can definitely feel the heat differential traveling from downtown to the suburbs 30 mins away by commuter rail. I’m sure this leads to more heat deaths in the city than in the surrounding area. I don’t mean to imply that these health effects aren’t significant. I just don’t know how to compare them to other factors, or even what factors they should be compared with.