I’ve lost track of where I saw the pointer to this page. Either at WUWT, or here hidden somewhere in “tips”. If you put up a tip, holler for a h/t…
I’d opened the window on it weeks? ago and didn’t have time to even read it. I ought to have. It’s very short, and very nicely done. It is something I’d started to do in a bulk way with GHCN (then found the max and min temps very lacking and let it go for “someday” that never quite came).
Basically, it compares what is happening to the MAX temps vs the MIN and MEAN (average) at each of a bunch of thermometers. No globally averaging homogenized who-haws… Just simple station data, made into series and compared.
It looks at the simple question of “Is it warming during the day, or during the night?”. The CO2 ‘back radiation’ thesis would have the evenings staying warm longer (‘holding in’ the daily heat a bit longer and better). The “sun did it” (or via clouds) thesis would have more energy arriving during the day, so days would warm more. As we’ve seen already, heat that arrives during a day, rapidly leaves, so the evening lows would tend to stay about the same as always.
So this is the game. We have reasonable accurate data from weather stations all over world for the past 35 years. I do not trust most of the data from long before 1975 unless it can be explained exactly how the measurements were done and recorded. (like in the example of the station in Armagh – Northern-Ireland).
So far, after evaluating 22 weather stations the score on Henry’s pool table for global warming is as follows:
MAXIMA: rising at a speed of 0.0382 degrees C per annum
MEANS : increasing at a speed of 0.0137 degrees C per annum
MINIMA: creeping up at 0.0056 degrees C per annum
HUMIDITY: decreasing at a rate of -0.02% RH per annum
PRECIPITATION: slight change at + 0.16 mm /month /year
The latest tables show that, over the past 4 decades, the rates of increase of temperatures on earth i.e. maxima, means (=average temperatures) and minima have risen at a ratio of about 7:3:1. Remember: these are the summaries of actual measured results from a number of weather stations all around the world….No junk science. No hypothesis. Every black figure on the tables is coming from a separate file of figures. Obviously I am able to provide these files of every black figure on the table.
Yes, it’s only a small number of stations that have all that data in all that detail available for this type of comparison. But, frankly, I’d rather have a few stations handled directly and cleanly than a thousand handled very badly (and that ‘badly’ version is exactly what GIStemp and HADCrut and NOAA/NCDC do – with all sorts of inexplicable prestidigitations, adjustments, homogenizing, and making up roughly 14,000 out of 16,000 ‘grid boxes’ in the present – as there are only 1200 ish currently active stations in the GHCN data set…)
So rather than homogenized ‘data food product’, here we get a smaller but much higher quality sample of Real Data.
And it shows that heat does not hang around all night long. It either soaks into some water somewhere or radiates away fairly promptly, such that the evenings cool back to ‘normal’, even when excess incoming daily heat puts a bit more in the system. (Henry leaves open the question of ‘why’ days are warmer, but suspects either more sunshine or fewer clouds. I think it’s both together. Hot sun and Svensmark lower clouds come together when they come).
Perhaps I’ll now be motivated enough to revisit the GHCN and see if the post adjustment manipulated data food product from NCDC has enough stations with mins and max to do a similar test. It would at least tell me if I need to just toss anything that has been through their hands and that we simply must restart from “Raw Data Only”. If a similar sample of stations shows the same min vs max trends, I’m confident we’ve “got something”. If they are suddenly nicely uniform in day vs night behaviour, well, that would be just too suspicious to ignore… There is no way days and nights will behave one way in the direct sample data and another in the post processed form unless the processing does something to the data.
For me, the approach is a simple, elegant, and direct one. It finds increased daily heating, and nights ‘not so much’. I think that matters…