surfacestations.org project. For those who don't know, the purpose of the project is to survey the condition of the surface weather stations that collect raw data which eventually gets adjusted and fed into various global warming computer models. The survey process includes photographing the stations, taking various measurements, and pinpointing their exact location. The results for each station surveyed so far are documented at surfacestations.org.
With 67% of the 1221 United States Historical Climatological Network (USHCN) stations surveyed, here's how the stations rate (using the the Climate Reference Network's objective Rating Guide):
* 11% have and expected error of >= 5 degrees C because their temperature sensor is located next to/above an artificial heating source, such a building, roof top, parking lot, or concrete surface.”
* 58% have and expected error of >= 2 degrees C because they are located within 10 meters of an artificial heat source.
So, almost 70% of the stations surveyed to date are rated as overstating the temperature by 2 degrees C or more. Note that this expected error is greater than than the entire increase in average global temperatures over the last century!
To be fair, the stations surveyed to date have been the more centrally located ones. As more rural stations are added, they will presumably be less likely to be located near artificial heat sources, and this will bring the above percentages down some. But if this is true, then one must adjust the raw data provided by these stations to account for the effect of urban sprawl over the last century: Presumably, a far greater percentage of stations today are located near artificial heat sources than a century ago. A century ago, the lot where my house sits was located in the middle of rural farmland. Today, it's smack dab in the middle of suburbia.