Urban Heat Islands and Optimal Density Levels

There have been subterranean ruminations in the comments sections of this blog regarding the recent column that urbanologist Joel Kotkin published Sunday in the Washington Post. Kotkin, as he always does, rose to the defense of the “suburbs” (a term he leaves undefined, but which presumably refers to those parts of metropolitan regions lying outside the traditional urban cores) in the context of Global Warming.

While I disagree with many of Kotkin’s conclusions, I always find him worth reading. He often raises points that force me to think through my own positions more carefully. For instance, commentators in the “smart growth” camp (or, in my case, the market-oriented sub-camp of the smart growth movement) contend that “sprawl” (by which, I presume, he refers to scattered, disconnected, low-density development) is wasteful of energy. As we remind readers endlessly in Bacon’s Rebellion, that’s because “sprawl” makes people drive greater distances and consume more gasoline.

But Kotkin responds that the world’s “cities” (by which he presumably means the high-density urban core of metro regions) are energy inefficient, too, in their own way.

Studies in cities around the world — Beijing, Rome, London, Tokyo, Los Angeles and more — have found that packed concentrations of concrete, asphalt, steel and glass can contribute to a phenomenon known as “heat islands” far more than typically low-density, tree-shaded suburban landscapes. As an October 2006 article in the New Scientist highlighted, “cities can be a couple of degrees warmer during the day and up to 6C [11 degrees Fahrenheit] warmer at night.” Recent studies out of Australia and Greece, as well as studies on U.S. cities, have also documented this difference in warming between highly concentrated central cities and their surrounding areas. …

Urban heat islands increase the need for air conditioning, which has alarming consequences for energy consumption in our cities. Since air conditioning systems themselves generate heat, this produces a vicious cycle.

This sounds totally plausible to me. And it’s definitely an argument against those who would use government coercion and social engineering to pack people into higher density environments. But I would raise two points.

First, the fact that urban centers are energy sinks does not negate the fact that sprawling suburbs energy inefficient. From an energy efficiency viewpoint, there may well be an optimum level of density between the two extremes. Kotkin points to the “village” scale of Reston as a positive example. Well, as Ed Risse frequently observes, if the one million-plus inhabitants Of Fairfax County lived in communities developed at Reston densities, two-thirds of the county would be open space today. A Reston-style Fairfax would offer the best of both worlds: fewer vehicle miles driven, without the heat-island effects of downtown Washington.

Second, no one really knows the “best” human settlement pattern. Even if we could figure it out for one point in time, the ideal would change with new technologies, shifting demographic patterns and changing energy prices. That’s why we need a decentralized, flexible system of land use based upon marketplace principles that allow individuals to optimize their own personal good. Here’s the trick: Market signals will lead us astray if they are distorted by government action at the behest of organized special interests. We need to provide a level playing field in which households and enterprises pay their location-variable costs.

There’s one more trick: Ideally, those location-variable costs should cover externalities imposed upon society such as pollution and the necessity of safeguarding Middle Eastern oil supplies. How you do that, I’m not sure. But in the abstract, that’s what we need to do.

Share this article


(comments below)


(comments below)


9 responses to “Urban Heat Islands and Optimal Density Levels”

  1. Anonymous Avatar

    Well said.

    I believe there is a balance which is neither all large lot or all highrise. And the tipping point probably changes over time. It probably changes a lot faster than we can afford to redesign our living spaces, and faster than they wear out.

    Therefore theree will always be a cadre of people who live at (what some others might think is) suboptimal efficiency. We should stop wringing our hands over it, and just do the best we can with what we can see is happening.

    The energy information agency has some tables that show total residential energy use for rural and urban households, and it turns out they are about equal, even considering that rural households travel farther.

    But, Urban areas have enormous external energy useage that comes in on top of the residential usage, exterior lighting for example. I don’t think there is any doubt that, overall, urban areas use far more energy.

    Since a lot of the difference is public use energy, how are we going to make that part of the “market”?


  2. E M Risse Avatar

    Jim Bacon:

    I love it!

    You are setting up next weeks column like a champ.


  3. Groveton Avatar

    I have worked in the Reston Town Center since 1999. It is a great place to work. I live about 8 miles away in Great Falls so the commute is pretty short too. If there were bike paths in Great Falls I could ride a bike to work.

    I am in LA today (center of dysfunctional settlement but loads of fun). I’ll try to put out a message about the pros and cons of the Reston Town Center (many more pros than cons).

    I am sure that EMR is right about what Fairfax County would look like if all the people lived in Reston-like densities. The problem is that they don’t. Hard to see how you’ll get the people out of their sub-divisions now. The English subsidized farms to keep the farmers outside London happy while prohibiting development of those farms.

    Maybe Fairfax County needs to start buying whole subdivisions and bulldozing them?

  4. E M Risse Avatar

    Thank you Groveton.

    I still owe you that column on how Great Falls can become a Balanced Village. It will be a while.

    Have a great trip.


  5. Lyle Solla-Yates Avatar
    Lyle Solla-Yates


    Well put. The urban heat island effect can be managed. For example, painting roofs white instead of black retains less heat, while planting trees can help cool an area. High density buildings can even incorporate trees, as green architect Bill McDonough proposes.
    Still, I agree that there is such a thing as too much density, and that a well regulated market is the best way to find what that is.


  6. Anonymous Avatar

    The urban heat island effect also works in winter, spring, and fall in a good way. So a better question is how do you mitigate it in the summer only without breaking the bank.


  7. Anonymous Avatar

    The heat island comes two ways, from insolation which making white (or green) roofs helps. But much of it is also just energy escaping in the form of heat which making white roofs doesn’t help. I think the heat island represents the energy loss from the city more that it represents the solar gain. In any case the solar gain is greater in summer, when it is needed least.

    I knew a fellow who lived on his boat, in fall he would paint the decks dark gray, in spring white. My house has a dark roof, which I can get away with because it is shaded with trees.

    One might think that the heat island in winter means less energy use because it makes the buildings easier to heat, but it is a small effect compared to what it takes to maintain proper ventilation in the buildings. Modern buildings already incorporate solar exposure in their heating and ventilation software. Heaters turn down on the sunnyside of the bulding before the sun comes on it and makes that side uncomfortable, and the air flow is directed towards the cool side to use tht heat as long as possible before it is exhausted. As the sun trvels around the building, the process is reversed.

  8. Larry Gross Avatar
    Larry Gross

    this is one of those deals where geography might “count”.

    A city with good prevailing winds probably is cooler than one that does not.

    For instance, take two equally size cities, each producing the same amount of heat – and put one say.. where Vancover BC is and put the other one where.. Mexico City is.

    It would be interesting to see a chart with comparative – per capita or per footprint size data.

    If Global Warming is real (and we just disagree on causes), wouldn’t we have a ton more cities involved in this effect?

    From a global warming perspective – have we actually calculated the net fiscal and environmental impacts of heat islands if they are accelerated in both numbers cities and intensities of existing cities?

  9. Anonymous Avatar

    I was once sailing along the coast in the fog with a friend. When I turned and entered the harbor spot on, he asked how I did it.

    The wind coming off the shore was warmer in front of the town, and it wasn’t even a major city. The traffic along the shore turned inland to go around the harbor. First it gets warmer, then it gets quieter. Finally the wave pattern will change in front of the entrance because the current changes there.

    The environment can tell you a lot, if you just listen.

    Maybe heat islands are only one part of the total system we need to look at, but it is good to hear Larry saying things like “net fiscal and environmental impacts”


Leave a Reply