UHI and its impact on urban temperature records is a matter of physics, not mathematics. The effect is so stark that it registers clearly even on car thermometers when driving from the open countryside into the city and vice versa. Detailed scientific studies to quantify the effect have been
performed, inter alia, in Winnipeg, Melbourne and Tehran.
That there are several “qualified” people who think that UHI has negligible
effect on yearly average temperatures over climatic time-scales may be
true, but their thoughts are irrelevant to the scientific question. There
are tons of people who hold forth on matters quite beyond their field of
competence–especially so in “climate science.” Wedded to ad hoc methods
of constructing anomaly series from scraps of data, they seem unaware of
direct, apples-to-apples comparisons beween validated century-long records
from urban and neighboring small-town stations. It is only by making
measurement location a fixed, rather than uncontrolled, variable that truly
scientific determinations of UHI discrepancies can be made.
Unfortunately, there is such a dearth of valid, century-long, small-town
records outside the USA, that truly indicative comparisons cannot be
made on a geographically representative basis. That’s what opens the field
to unphysical, statistical speculation wherein the properties of the
inadequate data base are conflated with those of the real world. Urban
signaturesd are spread thereby to vast unmeasured areas by kriging.
Some months ago a perceptiver commentator on WUWT posted the results of a
bona fide compilation of discrepancies between a geographically
representative samples of urban and non-urban annual averages in the USA.
It is the first latest graph found here:
s1188.photobucket.com/skygram
These results are corroborated by my own work (which goes considerably beyond simple comparisons that I will not get into here). Before plunging further ahead, I’d urge you to become acquainted with the only posting I’ve seen on the web that presents physically meaningful results on UHI on a quasi-continental scale. BEST deserves
disparagement for avoiding meaningful comparisons