No global warming for 17 years, 6 months

I’ve just plotted this months data. You can now go back to June 1996 (rather than last month when you could only go back to September 1996) and still get a flat trend line. So I think it’s 17 years 9 months without warming. (or have I got it wrong, I’m no expert in this sort of thing)

[snip – fake email address]

Dodgy Geezer says:
March 4, 2014 at 4:22 am
… In my day everyone in the medical profession would have had a smattering of Latin at least, and usually Greek…
—
Quidquid Latine dictum sit altum viditur.

In a recent exchange on CNN’s post-article blogging, I pointed out the lack of warming over an interval so long and profound that the IPCC felt obligated to devote an entire explanatory panel (Box 9.2 in chapter 9) of AR5 to it, and a devout member of the church of warming made the assertion that if one examined the temperature record over the last 30 years one would see that the climate has been warming at the torrid pace of some 0.25-0.4 C per decade. Presumably, of course, including the fact that it was flat over the last fifteen years (at least!) of that interval.
Out of sheer curiosity — since that implies that at least 0.75 C of warming happened since 1983 — I plotted this:
http://www.woodfortrees.org/plot/hadcrut4gl/from:1983/to:2013
where I omitted 2014 at the end only because it is too early to consider it a “year” and W4T doesn’t do month by month entry on its interval selector.
Imagine my surprise! The temperature change from 1983 to 2013 is negative, both for HADCRUT4 and (less dramatically) for GISS LOTI.
Obviously this is a profound cherrypick, but it was particularly amusing in the context of the CNN discussion and assertion of “30 years” as that precise cherrypick by the warmist intoxicated on his own sense of world-saving self-importance.
If one wishes not to pick cherries in a discussion that from the beginning has been all about cherrypicking, confirmation bias, and a complete lack of shame associated with both (see D’Arigio’s remark to the effect of having to pick cherries if you want to make cherry pie to the US Senate) one might instead plot:
http://www.woodfortrees.org/plot/hadcrut4gl/from:1945/to:2013
Note that from 1945 to the present there has been approximately 0.4C of warming, end to end, with variation both up and down in between. Crudely, it was nearly flat from 1945 to 1977, rises by around 0.4 C from 1977 to 1997, and is nearly flat from 1997 to the present (a classic Hurst-Kolmogorov pattern). In the single twenty year warming interval visible in this record, temperatures went up at a rate of 0.2/C. Everywhere else in the record temperatures are basically flat.
These are facts, pure and simple. They don’t cherrypick anything, they simply make direct observations from the recorded data. One can do an excellent job of fitting the data with three straight lines on the intervals indicated, and while one could quibble about the exact year to start or stop the interval endpoints — 1977 or 1980? 1997 or 2000? — they do not alter the pattern or the magnitude of the rise, only how it is distributed. The entire rise occurred in an interval so short that it is at the limit that the warmists themselves are asserting as “necessary” to establish a climate trend. They are at the boundary of being doubly hoist on their own petard — being forced to stop moving the goalposts lest them move them to where their own argument loses all force! By their own words, the warming observed across the 80s and 90s is not a trend.
By the data, too.
If one averages the warming from 1943 to the present, it is a paltry 0.4/7 = 0.057C/decade. If one averages the warming from 1973 to the present it is a larger 0.4/4 = 0.1C/decade (because the global temperature was flat, of course, in between). If one averages the warming from 1983 to the present it is accidentally slightly negative — bad choice — but if one shifts the endpoint to the carefully selected year 1985, at the very bottom of the 0.6C drop in HADCRUT4 from 1983 to 1985 (a cooling rate of 3C/decade — good thing two years aren’t a “trend”, eh?) one can manage to get a 0.6C rise over 27 years, or a sizzling 0.22 C/decade, still lower than the “official” estimate of climate sensitivity. If one more reasonably (but still highly cherrypickily) picks the smoothish bottom around the negative 0.2 C spike in HADCRUT4 — which is almost certainly pure artifact, of course, for all that it is larger than their presumably 0.15 C “error” — one is back to 0.4/2.7 or around .15C/decade.
These, too, are pure observations. The absolute worst case, most optimal cherrypick of a warming interval in HADCRUT4 confines all 0.4C to roughly 15 years and is no worse than ~0.27C/decade to all the significant digits that could possibly matter if one hand picks the end points. 15 years is, we have been told, insufficient to establish a climate trend, but all longer intervals reduce the warming rate rapidly to 0.1C/decade or less!
How much less, of course, depends on how long an interval you choose. All the warming happens in a single 15 year stretch and is flat on both sides of it, so have fun watching the mean drop as one increases the sample size.
What, in the end, does all of this mean? The answer is profoundly simple. It means precisely what the data indicates, nothing more, nothing less. We cannot explain the data with any model, simple or complex, at this time. It means that from 1943 to the present the world has had temperatures that average out to flatlined for 55 of the 70 years, and went up 0.4C over 15 years — basically a step function increase. This is what it is. We can speculate as to the cause of the step function increase — increased atmospheric CO_2, a coincidence of two unusually strong solar maxima, the particular phases of atmospheric and oceanic circulation leading to strong ENSOs, all of the above working together. We can make equally unfounded claims as to why the temperatures were flat from 1943 to 1977 and why they’ve been flat from 1997 to the present. (“Unfounded”, in this context, means that one can assert all sorts of hypotheses but cannot combine the hypotheses into a quantitatively accurate predictive model that can be gradually validated or falsified by agreement with future data, and so one’s assertions have no foundation in observational science at this time.)
What we cannot do is use models that have already been falsified by the data to make predictions of future warming that exceed the maximum warming rate observed in the most carefully cherrypicked interval in all of HADCRUT4 that is marginally long enough to be considered a “trend” according to the modellers themselves. At the moment, the “median” climate sensitivity claims in AR5 of around 2.7C by 2100 exceeds warming rate of 2.66 one would obtain by extrapolating the rate observed from 1985-ish to 2000-ish (if we confine it still further to 1985 to 1998 — as we probably could — we can make it still larger, but only at the expense of using an interval so short that not even a one of the high priests of warmism could justify calling it an extrapolible trend).
Anyway, I thought I’d share the grins with you. A negative anomaly from 1983! Who would have imagined! (And only very, very slightly positive if you go to 2014).
rgb
[Recommend we consider promoting this message, appended to the one following, to its own thread. ]

Anthony,

REPLY: I think we should take the high road, and not label them with either. While I disagree with Dr. Spencer’s use of ‘Climate Nazis’, I defend his right to say it. – Anthony

+1
Steven Mosher,
Thanks for the anomaly examples, I didn’t know about them.
Jeffrey Dean,
Thanks for demonstrating that the failure of the ADL has teeth.

Thanks Robert. I will remember that example next time somebody accuses me of cherry picking. (You want cherries, I’ll give you cherries buddy!) :>