Update and confirmation of 'Global warming stopped 16 years ago' aka 'the pause'

Chris G,
The source for the first chart is the CET record. If you don’t know that, you need to read the WUWT archives for about six months to get up to speed.
Now look at this chart. You will notice that ∆CO2 always follows ∆T. All this time you have been laboring under the assumption that ∆T is caused by ∆CO2. As you can see, it is exactly the opposite.
When your premise is wrong, your conclusions will necessarily be wrong.

rgbatduke has an excellent post @Oct 21 7:51
An ignorant layman like myself finds posts of that sort most interesting. Up until a short time ago I was unaware that modern weather forecasting is merely the broad application of statistical chance, just slightly better than rolling dice or throwing darts at a board.
It’s a bit better than that with weather (which is “short term climate”). Weather forecasts, given satellites overhead providing excellent data and a fair understanding of the immediate atmospheric physics that drives it, tend to be excellent one to three days out, decent a week out, and start to break down fairly quickly after that. I don’t know the exact statistics of the lagged correlation between prediction and reality — perhaps Anthony does, if he is still reading along with this thread as it is his thing — but at some point the forecast regresses to the statistical mean behavior for the time of year.
Climate (which is “long term weather”) attempts to predict the variation in this mean from year to year. Well, that’s not quite true, because the random noise year to year exceeds the systematic variation in the signal by an order of magnitude or so, and the mean itself isn’t the mean, it is some sort of running mean that averages over this noise. So they try to predict the variation in the coarse grained average over a suitable window extending back from the present that has to be long enough to reduce the undesirable short term noise, short enough that it doesn’t eliminate the secular variation they are trying to predict or make it move so slowly as to be useless on the timescale they are trying to predict. The woodsfortrees site lets you do this yourself — you can apply a variety of statistical processing to the actual record your very own self and see the results graphed in real time, including selecting windows to do the coarse grain average. I can’t remember if they let you select windowing functions — the most common one is a square function but an arguably better choice is an exponential or half-gaussian window that smoothly reduces the contribution of earlier years rather than cuts them off all at once.
Hence the problem with signal and noise. We don’t know which is which, for climate. There are doubtless parts of the signal that represent secular time variation of the mean that is not “random” or chaotic noise but is solidly entrenched, causally linked stuff with timescales ranging from hours to centuries. We even know some of these causal phenomena and have an idea of their timescales. Examples of long term contributors to this at least partially deterministic secular variation include the Sun (which is itself a climate system that we also cannot predict particularly well and which affects the climate on Earth multiple ways only some of which we have a good physical handle on), the global decadal atmospheric oscillations (which are tied to the ocean), the ocean itself (which is coupled back to the atmosphere), and heterodyning in annual oscillations (constructive interference, if you like, between year to year “random” variations which usually cancel (producing the mean behavior) but often do not, producing the extremes of the distribution.
Since those extremes have a “lifetime” and are nonlinear coupled to everything else, the variations of the climate about some fairly arbitrarily defined “mean” aren’t really Gaussian, yearly climate isn’t a set of independent and identically distributed samples, and most of the assumptions underlying ordinary statistics break down so that concepts such as “mean” and “variance” and “standard deviation” — all of which give us predictive capabilities for iid samples drawn from stationary distributions with bounded variance via the central limit theorem — are a lot less meaningful and predictive. Nicholas Nassim Taleb’s book “The Black Swan” goes over this, primarily in the context of economic markets where a “Black Swan Event” is a rare event with very high impact, such as the one day the market loses all of the money it made in the previous thirty years, confounding the expectations of those that were following the short term trend in the windowed mean and thinking it had actual predictive value.
The exact same thing could happen in the case of climate, which is also nonlinear and full of unknown chaotic causality and feedback loops. It could happen in either direction, even. Right now the trend — in a fairly narrow window of past behavior imperfectly known and with a growing experimental uncertainty — is general, slow warming over the last 150 or so years, so it is “safe” to bet on more general, slow warming using the past is a good predictor of the future model and completely ignoring underlying possible causes as being so complicated that they might as well be random noise. However, the climate could “suddenly” downtrend into the next glaciation on top of the elevated CO_2 — it has in the past when CO_2 levels were more than ten times what they are today (Ordovician-Silurian transition) and there is no way we can be certain that it cannot happen again because we don’t know why it happened then, only that it did. It could also “suddenly” uptrend into a “catastrophically” warmer state. These are the moral equivalents of a Black Swan market crash, confounding all of the short-term statistical numerology based on the (too) simple assumption that today will be mostly like yesterday, somewhat like today was a year ago, less so like it was a century ago, and not at all like it was 20,000 years ago.
This is my primary bitch with climate science. It isn’t that it isn’t trying, and using “decent” methodology as far as it goes as it tries. It is that there is IMO insufficient attention paid to the limits of our knowledge and how much of our “predictions” are really numerology in a tightly coupled non-Markovian chaotic multimode system with unknown couplings and dynamics galore and at least three or four distinct hydrodynamic systems of enormous complexity (two on the Earth and at least two on the Sun) contributing, where hydrodynamic systems — let alone magnetohydrodynamic systems — are the most complicated and difficult mathematics in the known Universe, so far, relevant to the real world. Unified field theory is simple in comparison.
A fair statement of our ignorance would significantly reduce the certainty of our predictions of continued warming, increase the expected possibility of cooling, allow for the possibility of either catastrophic warming or cooling, recognize that while there is probably damn-all we can do about either the trends or the extremes as we don’t really have a good enough grasp on the dynamics to understand the secular variations in climate or the primary drivers of the climate over the last 2000 years (inferred by proxy to the extent that we can, with increasingly large error bars the further back one goes) or even the last 200, it is probably wiser to not twist the tiger’s toe where and as we can avoid it because we do not know for certain how the nonlinear system will respond to toe twists.
Again, there are numerous examples of “what can it hurt” activities that in hindsight caused market crashes in “rolling good times” — electronics companies pretending they were banks and loaning dotcom startups equipment in lieu of money, savings and loan companies that ignored the risks of downturns that would wipe out their investment strategies, the ability to buy on margin that literally created money out of thin air in the form of kited checks that were always paid off by the next rise in the market — until the day they weren’t and all of the virtual money in the system evaporated like the illusion that it always was.
Perhaps CO_2 is the devil itself, the CAGW enthusiasts are right, we are on the edge of runaway positive feedback that will cause a Black Swan climate event that turns the Earth into Venus complete with boiling oceans (a la Hansen’s most extreme nonsense). You have to envision not a certainty, but a gradually decreasing set of probabilities of these futures as they become more extreme, because you don’t know that this won’t happen any more than Hansen knows that it will, we only know that it hasn’t happened yet or happened in its more extreme forms anywhere in the several billion year long climate history of the world, as best as we can deduce it from paleontological records. If CO_2 is the devil, perhaps we can ameliorate or reduce the probability of the more extreme events by at least trying to keep from driving it up to concentrations 2-3 times what it has been in the relatively recent past, as those extremes could have an unexpected nonlinear impact on the ecosystem (not just the climate, and not just good or bad, but unexpected).
Perhaps CO_2 is harmless to highly beneficial, perhaps global climate is completely insensitive to it and all of the warming and cooling of the past has been simply coincidental where it has corresponded to matching changes in CO_2, or rather perhaps temperature determines CO_2 level instead of the other way around. Perhaps the world is on the edge of returning to a (little or big) ice age, the state of increased glaciation for the next 80,000 or so years. Perhaps our climate is determined by tiny variations in the local environment of the part of the galaxy the Sun happens to be passing through, only we haven’t been able to observe solar and galactic state well enough, long enough, to be able to see this. And don’t be certain that we could see it — if 90% of the mass of the galaxy is missing and invisible because it doesn’t directly couple to the electromagnetic field, solar state could vary proportional to the amount of dark matter that the sun “eats” as it passes through clouds of it in the galaxy, altering its gravitation. The solar interior/core is a hotbed of nuclear interaction, and dark matter may couple to nuclear interactions enough to be slowed down and captured within the sun, yet not coupled to light so that it cannot be driven away the way ordinary charged matter is as the solar wind.
Our ignorance is really pretty profound, still. We have come a long way since the Enlightenment, and we have the tools to learn almost anything, given time, but we haven’t had the time to build and use all of the tools and make the observations required to work it all out. Until we know what dark matter and dark energy really are (which could still include “nothing”, because they might not exist at all but something more mundane could be causing the phenomena we use to infer their possible existence) we are missing a pretty enormous piece of the physics puzzle and this has to reduce our confidence in all predictions of macroscopic scale phenomena like solar state and planetary climate.
In the meantime, it is pointless to panic and wreck the world’s economy and create a certain catastrophe to try to avoid an uncertain (and probably unlikely) catastrophe predicted by the more extreme variants of largely unsuccessful models. At the same time it is perfectly reasonable to take small steps to try to avoid twisting the toe of the tiger, and try to gradually convert our civilization away from carbon based energy. There are good reasons to do that anyway, and hedging one’s bets is only common sense (especially for Black Swan events). Taleb points out that ignoring the possibility of the extreme events and failing to be sufficiently conservative and hedge bets, even at the expense of making less money than the next guy, is what dooms long term investment first that make short term assumptions about the progress of the market. Perhaps humans should hedge the climate bet, instead of allowing our pockets to be picked by people who seek to exploit the sensationalism and create a panic that will make them rich or going around claiming that we are “certain” that CO_2 is a harmless trace gas, that the greenhouse effect doesn’t even exist, that there is zero chance of catastrophic warming, that an ice age is coming, that skeptics know what will happen to the climate ten years from now any more than the warmists do.
I’m unashamed to say that I don’t. In a decade, it might well be warmer. It might be catastrophically warming, the arctic melted, greenland melting, the antarctic finally melting to catch up, and all sorts of other bad stuff. It might well be pretty much the same. It might be cooler. It might even be a lot cooler, falling back from the current temperature peak to levels closer to those last seen in the 1960s, 1920s, or 1800s. If I had to pick, I’d pick the same to a bit cooler (betting on the sun trumping CO_2, basically), but I wouldn’t bet hundreds of dollars on it. I certainly wouldn’t bet hundreds of dollars on much warmer, which is what CAGW enthusiasts would have us all do. Nor would I bet hundreds of dollars that we will be “plunging back to a little ice age” even if we enter a Maunder minimum, although this is certainly possible. The extremes are generally worse bets than the middle; they just have much greater (more “catastrophic”) costs if they come home.
rgb