China's Urban Heat Island Problem

For some months I have been chipping away at writing a paper dealing with UHI. Most historic data gives us mainly the daily Tmax and Tmin, recorded when thermometers designed for the purpose cause maker pegs to stop moving each day when the max or min is reached. For 100 years, that is about all we have to work with.
Now, Tmax and Tmin are rather special temperatures, because they are reached when an interplay of thermal effects reaches a described point. At Tmax, for example, there has been a steady T rise as the sun moves higher in the sky, the rise helped by convection of air with hot packets in it surrounding the site, held back if frost has formed overnight, complicated if there is snow around and water phase change effects need consideration, hindered or lagged by the thermal inertia of the screen surrounding the thermometer as the screen heats up. There can be interruptions when low cloud causes lower T, there can be nearby vegetation that casts a shadow on the screen sometimes, other vegetation effects like when mowed surrounding grass changes the effective height above ground of the thermometer, there can be a burst of rain that cools the surroundings – and so on into the night. The actual Tmax recorded each day is thus subject to some weather effects and some non-weather effects, acting in a way that was not observed and recorded at the time and so lost to us forever.
Tmin has a similar story, with not much in common with Tmax once the sun sets. There is only scientific nonsense in subtracting Tmin from Tmax to give a diurnal temperature range. There is also nonsense in taking their arithmetic average and calling it a mean daily temperature, because the factors that cause Tmax do not have much in common with the Tmin factors and we do not now know the past details of each.
When it comes to UHI, where typically comparisons are made between urban and rural settings, there is again nonsense in interpretations because the factors that caused a certain Tmax in the city need not be acting in a similar way at the rural setting. Sure, one can use a blunt object approach and say look, the measured T is higher in the city than its burbs, so there is a UHI effect. The success of this approach has been poor, with people resorting to measuring night light seen from above as an index of a variable affecting T, or using population density maps likewise, without systematic study of whether each perturbing effect is real and quantifiable. IOW, just guesswork.
Guesswork becomes humour when the effects are trivial, but here we have $$$billions at stake for remediation proposals and that is not funny. It is destructive of Science at a rudimentary level with high cost outcomes. Much pain, no gain.
A colleague and friend, Dr Bill Johnston, has been working for several years with data that relate Tmax and Tmin to rainfall. Rain cools. Years with higher rain will typically show lower Tmax than dry years, Up to 60% of the variation in T data over time can be statistically explained by brainfall. It can be used to find breaks in time series T data and to show what homogenization attempts have been up to. For some major Australian sites like Melbourne Regional and Sydney Observatory, Bill has given good evidence that there has been no temperature change since commencement about 1860, when you delete the effects of cooling rain.
No temperature change, no global warming. Arrived at through thought and observation of the original data and a life spent in relevant scientific fields.
If Bill is correct, UHI is a measurement artefact, as is global warming. Think about it. Like, do cities modify the rainfall effect on Tmax in different ways to rural places? Can anyone answer that? Geoff.