James Hansen's group and colleagues at the Goddard Institute of Space Science (GISS) take the global temperatures and process them, correct them and out put maps and charts showing how the earth is warming incredibly and implying that we are all going to die before it is over (which may be true regardless of what happens to CO2--so the preachers tell me)
Anyway, sometimes it is really instructive to look at the corrections, to see what patterns they have geographically. To that end, I went to this site and made a map using the Hadley/Reyn Ocean data plus the unadjusted land data from 1880-2008 data. Next the gridded data was downloaded. I then made the same map using the GISS adjusted data and downloaded it. Then, after subtracting one grid from another, one can see what the magnitude of the adjustments are. Then I averaged the difference over both the latitudinal and longitudinal axes producing two graphs. Just so you can visually see the difference between the two maps, here is the unadjusted data followed by the adjusted map.
Note that the adjusted map has hotter temperatures. This is a very important point: a scientist only corrects the data if he thinks he knows it is bad and how much it is bad. In the case of thermometers, this means that they think the thermometers they correct are not working properly. That says that in general GISS has to think that the raw data is reading far too cold and must have their readings pushed upwards. There really is no other explanation for the additional heat. We will apply this logic to their corrections.
So, now that we have seen the maps, let's look at the latitudinal distribution of the corrections. This is very interesting.
When all is boiled down to its essence, the GISS temperature editors must be believing that southern hemisphere thermometers work just fine. They need no corrections, or very small corrections. But WOAH, move that thermometer into the northern hemisphere and it goes cold on them. How interesting that the laws of physics change between the hemispheres. Metallic expansion, upon which most thermometers are based, must behave differently in the north than in the south because the GISS folks feel the need to correct the north far in excess of the correction they give to the southern hemisphere.
Of course, the conclusions drawn from their corrections are silly. But that doesn't mean the logic is faulty, it means that the corrections are silly. Why would modern thermometers in, say Argentina or South Africa work better than those in Europe? They shouldn't.
Let's look at the longitudinal distribution of their adjustments. In this case I averaged the adjustments over the longitude bins.
Woah!!! This says that the GISS climatologists somehow believe that a thermometer in Africa or Europe needs as much as 8x more correction than a thermometer in China. Wow. What a discovery. And since the only way that the max correction can be both in the northern hemisphere and between 0 and 81 deg longitude is for the bad thermometers to be in Europe. All of the European thermometers read too cold and must have heat added to them. So much for German engineering. What happens to a thermometer in Europe that makes the GISS climatologists think that they are all reading too cold? Are all the thermometers inside shopping malls? I doubt it. But since GISS won't change the temperature of a thermometer unless there is something wrong with it, clearly they think that European thermometers are just crap.
The logical conclusion from all this? If you want your thermometer to work well, don't move it to Europe, Africa or the Northern Hemisphere. Carry on.