Examining Urban Heat Islands – Part 2

A look at how urban areas might cool

Part 1 looked at the growth of various cities and urban areas and how that might cause UHI in the temperature record, which is likely to be a tangled mess of types of stations that all need to be treated slightly differently. How BEST to deal with cooling cities started to look at this issue but there’s plenty more to say.

BEST finds no statistically significant UHI effect but yet contradicts findings with words (see Muller’s problem with UHI, for example). The biggest issue is that there’s a real mix of urban stations – some warming and some cooling, and the same with rural stations. We know from the BEST dataset that ~30% of stations show cooling.  Of course both NCDC and GISS use comparisons with surrounding stations to correct for urban warming – but if the surrounding ‘rural’ stations, which may be affected by land use change or rapid growth, are warming while the city is not this ‘rural’ warming trend is imparted to the city record also. The inherent thinking in their adjustments seems to  be “Why would cities cool? This must be wrong.”

Oke (1983) suggests  causes for UHI (Table 2, pg17), including “decreased evapotranspiration” due to reduced vegetation and an increase of impermeable surfaces.  Rainfall runs off quickly and doesn’t hang around to increase evaporation.  Even if there is minimal surface water, vegetation mobilises water from the subsurface and transpires it into the air as a means of self-regulating temperature. Put simply – vegetation cools. Apparently transpiration accounts for about 10 percent of the moisture in the atmosphere, with evaporation from bodies of water making up most of the other 90%.   As the following set of maps show, the darker areas on the left hand image (areas with more vegetation) are cooler both by day and night. You can also see a reservoir, located at about 7 o’clock in each image, is cool during the day but warm at night.

Source of images: http://envstudies.brown.edu/reports/TreeReportForWebPrime.pdf

Increased greenery was being promoted by the source report for this figure as a way of cooling the city and mitigating the effects of climate change.  Green roofs are another way – just don’t use them for siting weather stations – Thermal absorption – a black and green issue.

In arid areas, development brings vegetation and irrigation – take a look at the difference to Las Vegas in 25 years of growth – it has, literally, been blooming in the desert.

False-colour Landsat images of Las Vegas taken in 1984 and 2009 showing the increase in urban sprawl. [Vegetation – green, city – grey] Source: http://edition.cnn.com/SPECIALS/world/road-to-rio/satellite-photos-urban-sprawl/index.html

Changes in albedo are important too. The airport, which is the WMO station, can be seen as dark in the desert on the left but pale against the city (centre right), and look at all those irrigated lawns, parks and golf courses. Las Vegas, however, is not cooling – more about this in Part 3.

Young towns and cities grow…

A developer clears a patch of land of trees shrubs and tall grass and weeds. Roads and utilities are put in as housing plots are laid out.  Houses are built, people move in. Neat lawns are sown, gardens are planted.  Urban sprawl develops. We can imagine this land use change and infrastructure development would contribute to UHI, warming the city.  As we saw in Part 1, growth of towns and cities has been rapid at some times, then slows down.  How might this affect UHI and what happens when growth slows?

Site maturity

As I said “…lawns are sown, gardens are planted…” well, shrubs and trees grow. Typically they mature in 20-30 years – here’s the effect:

Suburb of Madison, Wisconsin (Image from Google Maps)

The green circled area on the left is ‘young housing’ with grass and the beginnings of gardens. The yellow area shows well-developed shrubs and developing trees, while the red area (top) has mature trees, providing shade. The change overall increases transpiration again.  I keep thinking of all that rapid growth of towns and cities during the Baby Boomer years in the 50s. If this greening is what happens to suburban areas over 30 years, does this mitigate warming and even allow cooling? This IMHO is worth investigating, however challenging.

Overall this possibility means we can’t generalise and we can’t use population increase alone to correlate UHI, unless tree growth is ruled out. Once again we only get estimates. Perhaps the greening only offsets other urban development. The effect will be entirely different in warm/cool areas and dry/temperate/wet areas. It is likely in some areas, but here, again, is one of those issues that probably requires a lot of attention to detail of metadata – correct metadata.  If there is an effect from the growth of ‘urban forest’ then we need to be careful WHEN we apply UHI corrections. How do we do that worldwide for historic data?

Chasing down UHI starts to become even more like looking for a needle in a haystack and it is easy to see why it is more convenient to say it doesn’t exit.

“I think China is one of the few places that are affected [by UHI]. The paper shows that London and Vienna (and also New York) are not affected in the 20th century.”
Phil Jones – email 1601.txt