Europe has a whole lot of data from a very long period of time. I had to make this graph longer and with wider range to hold it and be readable, so it’s really ‘squashed’ in the thumbnail seen here. Open the graph for something more easily readable.
The most striking thing, to me, is just that the v3 data look to be ‘offset’ by about 1/2 C warmer over much of the history of the data set. This will tend to reduce any ‘trend’ found. Then again, with 1.5 C or more of “warming” from the mid 1800′s to 1990, perhaps it was starting to look a bit silly (IF the 1990 to 2012 data were to show even just 1/2 C of added ‘warming’ you start to reach incredible 2 C and more warming in just over 100 years). Perhaps, as well, it lets these “Plus” adjustments offset “minus” adjustments in some other part of the data. Most likely it’s just FUBAR.
Looking at the dT lines (light blue and yellow) it looks like v3 is more volatile, but with a bit more ‘volatile high’ than volatile low. There is a fairly pronounced 50 to 60 year “ripple” with highs in the 1880s and 1930s (so now we ought to be cooling and 1990 ought to have been a peak… always nice to end your data series on a cyclical peak if you want to find a bogus rising trend…) and we also have both “Eighteen hundred and froze to death” and whatever the cold excursion was in the mid 1830s showing up. Then we rise to a roughly -1/2 C / -1.5 C range by about 1860 that holds, modulo the 50-60 year ripple, until the early 1980s (though with a hot mid 1930s as we’ve seen in the records set then).
Finally, between 1987 and 1990 we get that added 1/2 C “offset” bump that we’ve seen in other series. As the ASOS were rolled out to airports, and as stations on an RS-232 cable were pulled closer to buildings, and as the humidity sensing automated systems that were “sucking their own exhaust” added artificial station heating. There was also a change of “Modfication Flag” or “Duplicate Number” in the data records then. IMHO, this is just more confirmation that there’s 1/2 C of “warming” from that change in those few years. 1987 to 1991 as a step function. It is just NOT a long slow CO2 driven rise.
That’s it for a while. I’m going to do “other stuff” for the rest of the weekend. (Like catch up on some financial work / postings). Some time next week I’ll take a look into individual countries and see if there’s a deeper pattern. What I see out of these ‘by Region / Continent’ graphs is just that there are widely divergent behaviours from region to region; something NOT consistent with a widely mixed gas. There are local impacts and changes, not consistent global impacts.
Overall, the result just looks very “artifact rich” and very “physics poor”. IMHO, the temperature record is not “fit for purpose” in any “climate science” analysis that depends on greater than 1 or 2 C of precision.