Don't have research to back this up, but I believe some who challenge global warming do so on the basis that urban heat islands have skewed our temperature records, creating a spurious rise in temperature.
Now I could agree that a heat island effect could skew the record of maximum temperatures at a given location. If an airport was mostly rural back in the day, but now is in a urban area, the maximum temperatures are likely to be higher than otherwise.
But that's not an argument against global warming, just an argument against reliance using record high temps at a site as evidence for it. I'm assuming that the experts create an average temperature for a given area by taking the temperature at a point and applying it to the surface area around, extending the area until it reaches the area represented by another point. For example, take Dulles airport, which is maybe 6 miles west of Reston. If Dulles is at 80 degrees today and Reston is at 78, then in my mind the average temp would be 79, as Dulles represents the area approximately 3 miles east of the airport and Reston the area 3 miles west. Determine how many square miles Dulles represents and multiply times its temp, do the same for Reston, and all the other weather stations and you can come up with an average temperature.
If this image is right, see what it does for heat islands. The heat in an island is real, so if Dulles gets built up so it's now 82 degrees rather than 80, the 82 should still be applied to the area around Dulles.