Yes, Urbanization Does Put an Upward Bias on the Surface Temperature Record
This is one of those issues that really should surprise no one, but encroaching urbanization on surface temperature measurement stations can impose an upward bias to recorded temperatures, creating a false trend. The increase in measured temperatures due to urbanization is easy to demonstrate -- my son and I did it as a junior high science project.
The NOAA has a paper out that confirms the effect on surface temperature measurement. By the way the UofA temperature station photo illustrating the photo was actually taken by yours truly, becoming the most circulated photo I have ever taken. Here is the story.
In short, what happens is this. Urban environments are hotter than the surrounding countryside, so temperatures in the city will be biased upwards from those in the country around it (you will often see this on the local weather when they contrast the city vs outlying areas). This in and of itself does not necessarily corrupt the temperature trend. However, if the city is growing -- say in the case of the UofA photo in the article which 100 years ago was in an huge open field -- then encroaching urbanization can bias the trend.
Even with these biases removed, it is important to note that there is still an upward trend in the surface temperature record, at least over the last 30 years (as there is in satellite temperature measurement which is not subject to this bias). However, the total US surface trend may be overstated by a third to a half. Climate scientists of the alarmist sort have one of two reactions to this: 1) There are urban heat island deniers, who deny it is an issue or has any effect on the temperature record; and 2) There are those who accept that it exists but claim it is accounted for by various statistical methods that look at multiple sites in one area. The problem with this latter is that rather than actually remove the bias, it tends to smooth the bias like peanut butter across multiple stations.
