And here's the kicker: remember, the whole point of the paper is to show there is no difference in temperature trends between
specifically rural sites and the unknown mix of gridded networks, and then use this to claim that any UHI effect is very small.