Global Warming Canards: Urban Heat Islands

Since the 1850's we have had thermometer readings from enough areas around the earth to get a good idea of global temperatures. From those readings we can determine that the earth has undergone unprecedented heating. However, one canard used by global warming denialists says that isn't the case. This canard says that during that same time period we have undergone unprecedented urbanization. Cities are warmer than the surrounding countryside. Thus, the warming we are seeing is not the result of greenhouse gas emission but simply an artifact of recording from urban centers.

The principle this canard is based on is called the Urban Heat Island (UHI) effect. In order to do so, we need to understand the basis of heating and cooling. Heat is a measure of energy within a system. The more energy one puts into a system, the warmer it is. For the purposes of understanding the UHI effect the two systems we will look at are adjacent rural and urban areas. The temperature of either system will be dependent upon the energy input minus the energy loss.

Fig 1. Energy balance of rural and urban areas

This figure shows an example of the energy balance between an urban area (Houston, TX) and the surrounding rural area. If we do the math we get:

ENERGY INPUT_rural is 13.8 Kilowatt hours per square meter per day.
ENERGY LOSS_rural is 13.2 Kilowatt hours per square meter per day.

ENERGY INPUT_urban is 15.3 Kilowatt hours per square meter per day.
ENERGY LOSS_urban is 12.7 Kilowatt hours per square meter per day.

The balance sheet then shows that for a rural area there is 0.6 Kilowatt hours per square meter per day energy that can go into creating heat, while in the city there is 2.6. Thus, the city can be quite hotter than the surrounding countryside.

Fig 2. Temperature profile of the city and surrounding countryside

As this figure shows, the difference can be dramatic. If all the thermometer readings were taken from central downtown areas then the global warming data would be very skewed so much so that globe would have had to undergo significant cooling to produce it. So, as with most global warming canards there is an air of plausibility to it. However, the UHI effect was first noted in 1820. It would be surprising if climatologists were unaware of it. Well, it turns out we have no reason to be surprised. Climatologists ARE aware of the potential problem. And like all scientists when faced with a problem they investigate it.

The first step is to form an hypothesis: Urban Heat Islands are skewing the temperature readings. I'll call this hypothesis UHI(skew). The next step is to determine predictions the hypothesis makes.

The first prediction is that temperature measurments from other sources would show less heating than surface temperature readings. They don't. Satellite readings, estimates from bore holes, estimates from sea floors all give the same trend.

Not all temperature recording sites are found in urban areas. Many are rural. The UHI(skew) hypothesis predicts that warming should be more pronounced from urban sites than from rural sites. What do we find? It turns out that long-term warming from 1880 to 1998 was slightly higher in rural areas than in urban areas (0.7^o C in rural areas vs 0.65^o C in urban areas). This contradicts the hypothesis.

Fig. 2 UHI temperature profile

This figure shows the UHI effect at night and during the day. Notice that the difference in temperature between the urban area and the rural area is greater during the night. The reason for this is that during the night is when the earth radiates heat. In rural areas this radiant heat is lost to outer space. In urban areas a lot of it is absorbed by other buildings and retained. Thus, the hypothesis that the UHI effect is skewing the data makes the prediction that the skewing should be more noticeable when comparing night temperature readings to day temperature readings. That data is readily available. When that prediction is tested we find ... nothing. There was no significant difference.

The UHI(skew) hypothesis predicts that the largest temperature rises should be in the most populous latitudes. It isn't. The largest temperature rises have been seen in arctic regions. Another failed prediction.

Finally, the UHI(skew) hypothesis predicts that if global warming is an artifact of Urban Heat Islands, then we shouldn't see it in data concerning sea surface temperatures. There is no UHI effect there. However, sea surface temperature readings strongly show global warming

So we have five independent tests of UHI(skew). All of them falsify the hypothesis. That in itself is pretty strong. But in science we can't stop there. There is one more question to ask. Why isn't the UHI effect skewing the data? One reason is that the recording stations are very heterogeneous. Even the ones from urban areas are often in cool spaces. For instance many are at airports lying on the far outskirts of town. Others are near parks. Another reason is that the few recording sites (less than 1% of the total) that are located in areas in which the UHI is prominent are excluded from the data.

So in summary, Urban Heat Islands have a local effect on temperatures, but this effect does not contaminate the data used by climatologists to assess global warming. It's effect has been measured, compensated for, and does not explain any of the pattern of global warming that the data shows.