Another “escalator“

Even better, the skepticalscience graph starts in 1973– the lowest point of the last cooling event we had. That’s how AGW fetishists see the climate– in chunks small enough to support their pre-selected conclusions.

The first graph looks like/resembles the graph that Bob tisdale presented to show zero AGW because the steps coincided with the El Nino events. I prefer Bob’s explanation.

Running your plot up to the end of data looks like about the most convincing argument for a strong CO2 effect I have ever seen. Quite worrying.
Ah, but don’t stop there. Why “550 months”? 600 months is a fifty year centered average, but it is pretty arbitrary, right? We don’t know the time constants for any of the processes associated with the hypothetical oceanic turnover, and of course it isn’t going to show any of the recent flattening of temperature or the effect of the last two solar cycles as they simply haven’t been around long enough to show up. Of course if you drop it to 60 months (five year centered average) nothing interesting shows up. If you go to 800 months, you don’t HAVE to clip the ends, they get clipped for you. The “perfect” correspondence goes away, of course, but you still get that correlation that is still (last time I looked) not causality.
And note well, Lief hasn’t even shown up yet to point out that the SSN counts in this database are all incorrect pre-1980, and that by the time they are all corrected even the 550 month truncated correspondence is largely gone.
Let’s not forget, others would have us plotting planetary influences against the temperature, so even if any given single parameter fits fails it doesn’t mean that it is wrong, only that other things are right as well. You assert that it is convincing to you that CO_2 might be the cause of the divergence at the end, but if you plot the same temperature series against carbon dioxide, you get an absurdly inadequate result because CO_2 was almost flat over the first 2/3 to 3/4 of the graph and has been a smooth monotonic function throughout where the temperature curve has a variety of slope and curvature variations.
And then there are the global decadal oscillations: ENSO, the PDO, NAO. It actually isn’t completely crazy that these oscillations might substantively moderate the overall heating/cooling efficiency of the planet on a decadal timescale, given that they make significant changes in where heat picked up one place is transported to another, lifted up to be radiated away, or modulates cloud cover and hence albedo or cloud/water vapor GHE (where there is 30 times more water vapor in the troposphere at any given time than CO_2, and where it is a potent GHG all by itself responsible for the vast majority of the GHE). Bob Tisdale has argued persuasively that ENSO alone is largely responsible for the “escalator” pattern in the late 20th century temperature series, not CO_2.
Somewhere under there are some very slowly varying causes — variations in orbital cycle, for example — but those causes are nonlinearly coupled to everything else so it is not really safe to say that they are irrelevant. In the long-time-series historical record, they very likely constitute a nearly irresistible factor that drives some sort of set-point for the climate, quite possibly with a decadal to century timescale lag, so the Earth could still be “catching up” to variations that occurred 600 years ago and that are responsible for some unknown fraction of the curve.
At some point, plotting “global T” against any single parameter, especially a monotonic parameter over an interval where “global T” itself was nearly monotonic, becomes a completely empty exercise. That point was reached long ago. It is perfectly clear that no single driver can explain the global temperature series of the entire Pliestocene. We have no single-parameter model that can describe the temperature series for the last billion years. We have no model that can describe the temperature series (deduced from e.g. ice cores) for the Holocene. We have no model that can describe the last 2000 years. We have no model that can describe the last 1000 years. We have no model that can describe the last 500 years. Not single parameter, not two parameter, not ten parameter (well, we can fit anything with ten parameters, of course, but at that point we can fit it hundreds of ways and none of the extrapolate or hindcast the rest of the series above and hence all of them are meaningless.
Can I point out one more time that this is a hard problem? One that is unlikely to yield to mere numerology?
And don’t even get me started on the data itself, or the fact that nobody makes woodsfortrees plots with error bars. I mean if one were going to try to fit the data, or look at any sort of correlation matrix or covariance matrix of the two (or more) proposed correlated entities, that would be one thing, but the significance of the fit depends pretty strongly on the error bars one assigns to the points on the basis of unknown factors, since none of the data represented there is pulled from an iid process. One of the factors is data “adjustment”, for example, where a stunning series of adjustments have almost universally increased the perceived warming in the temperature series.
IMO, one can really only trust the satellite derived post 1979 temperature estimates. Before that time, it is too, too easy to tweak the data within very large error bars and accept this, reject that, UHI another bit, and end up (somehow) with a lot more warming than one might expect looking at the raw thermometry. How one can tweak a UHI to make the series warmer in the present is a good trick, of course.
Finally, one needs to read about Hurst-Kolmogorov in order to understand the data escalator. This works both ways — the current flat is indeed not necessarily significant. Neither is the trend it is superimposed on. Insufficient information in a strongly coupled multivariate nonlinear system.
rgb

That first chart is not how I view “global warming”, so maybe I’m not a skeptic. Then again, I AM a realist, and the big red line that slashes its way through the clearly visible 180 degrees of a sine wave is highly offensive.
On the third hand, the “skeptical science” site has very little to do with skeptics, so… hmm.
I still wonder how “climate change panickers” are going to handle the next 10 years, when their beloved (but still fictional and meaningless) “global average temperature” starts dropping farther away from even the most optimistic “best case” scenarios. No matter how you slice it, more warming is off the table for the next few decades. And that will eventually be impossible to avoid.

First I’d say Leif Svalgaard wouldn’t approve that sunspot graph.
And second, averaging over 44 years??? Come on. Hasn’t here been an article just recently about how unethical it is to use smoothing to improve your correlation?

I would suggest doing a competing graphic. Show “how skeptics view global warming” and show the last 1200 years of temperatures. It’s a downward slope. Then show “how alarmists view global warming” where the the 1200 to 40 years ago record gradually disappears, leaving them with their exact graphic doing it’s comparisons as above.

Forget Leif. Hell my DOG thinks this is crap not worth investigating!