2. 1 Trees & Temps (Durham)

Figure 2.1: These images show distributions of Durham, North Carolina, tree cover in 2005 (30 m square pixels; darker areas depict higher canopy cover) and midmorning temperatures on May 8, 2005 (120 m square pixels; darker areas represent lower temperatures). The warm central core, about 5 km wide, marks the city of Durham. Lakes show up as regions with low canopy cover and low temperatures (see Figure A.2). The bottom image zooms in on Southpoint Mall, with dark areas being warm and small dots indicating impervious surfaces. (All images courtesy of Joe Sexton.)

Impervious surfaces come with urbanization, surfaces that rainwater can’t infiltrate, and include anything other than ground cover: buildings, parking lots, streets, sidewalks, bike paths, and garden sheds.[1] With these surfaces come urban heating. and this chapter examines, in part, the resulting “urban heat island.”[2] This wonderful example (Figure 2.1) from Durham, North Carolina, demonstrates that high temperature comes with low vegetation, with a nearly 5 km diameter vegetatively depauperated, high temperature urban core.[3]

Where is this urban heat coming from? I’ve often wondered whether importing fossil fuels and electrical energy into the city and using it there might contribute to the urban heat island: Think of a toaster with its coils glowing orange in the city, heating up the space within it, but the cord and plug-in leading off somewhere to a coal-fired power plant off in the rural hinterlands. Is human energy use important to the heat island? As the calculations connected with Figure 3.17 demonstrate, an average North Carolinian uses about 83 Btu/m2 of energy each day. In any event, this is a trivial amount compared with the Sun’s energy input of about 13,600 Btu/m2/day. Human energy input seems pretty minimal, at least for Durham.[4]

This comparison means that the heat in the urban heat island comes from the Sun, and the heat hangs around because impervious surfaces in urban areas reflect and absorb sunlight differently from rural areas. However, not all impervious surfaces are created thermally equal. I have a shed with a clear roof made of polycarbonate plastic: It has no thermal mass to speak of, but it intercepts rain just like my concrete driveway. Rain falling on both surfaces ultimately reaches the ground in my yard rather than going directly into the stormwater system, but the unshaded fraction of concrete also soaks up solar energy during the day and releases it over a long period of time. Asphalt shingles on my house behave somewhere in between these extreme thermal mass examples.

The bottom image zooms in on Southpoint mall in southern Durham County (see the aerial view in Figure 5.7), showing several fascinating points. First, the mall itself, with all its asphalt, is quite warm. Second, neighborhoods see much variation in their heat island fingerprints. Finally, New Hope Creek, which flows into the Jordan Lake reservoir, sits on the left side and shows very cool temperatures. Along this creek bed sit high impervious surface developments, and sharp thermal boundaries are visible between the creek and developments.


[1]When rain hits an impervious surface, it doesn’t directly seep into the ground. That elsewhere might be the side of the road, alongside a building, into rain barrels, or down a stormwater pipe.

[2]An excellent overview of the urban heat island is given by Pickett et al. (2001). The first published mention of urban heat islands, at least that I’m aware of, comes from temperatures measured in and around London by Luke Howard, in an 1833 book titled, The Climate of London, Deduced from Meteorological Observations, Made in the Metropolis, and at Various Places around It. These temperatures, taken from 1807 to 1816, show an urban temperature of a degree or two Farhenheit higher than the surrounding countryside.

[3]Canopy and temperature maps of Durham, North Carolina, were provided by Joe Sexton using the magic of geographical information system (GIS) software. The canopy map uses 30 m square pixels, and the temperature map uses 120 m square pixels. Color versions of these images are available and provide more detailed representations. Be aware that urban heat islands don’t happen every day. Thermal images from fall, winter, and early spring show absolutely no sign of the city, and even a few days of wet, overcast summer weather leaves no evidence of an urban heat island.

[4]An extensive review on the urban heat island by Arnfield (2003, p. 7) cites numbers ranging from 11 to 1,590 W/m2 for human energy input. The highest level was for Tokyo in the wintertime. My rough, county-level-averaged number of 83 Btu/m2/day=1W/m2 is an order of magnitude lower than this range’s lower bound. My energy calculation makes the fundamental assumption that people are spread out uniformly, but in city cores maybe this assumption doesn’t hold true. The City of Durham has three times the density of the county, and the city core is even denser; that probably brings my number up to the lower levels of this range. Even so, human inputs in Durham are small, maybe 5%, compared to solar radiation input of 13,600 Btu/m2/day, or around 160 W/m2. Perhaps having 100 times more people than average in a tall office building results in energy use heat inputs 100 times greater, which multiplying out gives 8,000 Btu/m2/day, approaching that of solar input. Interesting plots might be energy use versus temperature, and population density versus temperature, though correlations between energy use and industrial/commercial land use would likely get in the way.

One Response to “2. 1 Trees & Temps (Durham)”

  1. Will Wilson says:

    A paper by Oke (1988) reports anthropogenic heat inputs ranging from 6 W/m2 (Anchorage, AK) to 159 W/m2 (Manhattan). Still much higher than my estimated value.

    On further reflection, consider Manhattan at 159 W/m2. The conversion is 1W=3.4 Btu/hr, making it 542 Btu/m2/hr, or about 13,000 Btu/m2/day. Elsewhere I mention solar energy at Earth’s surface being around 16,200 Btu/m2/day. Kiehl and Trenberth (1997) reports this solar input as 168 W/m2 at Earth’s surface. Hmmm. Human input is pretty close, or sometimes greater.