cohenite, I fear you have little understanding of statistics. Let me break this to you gently, just because a statistical analysis of a set of numbers divorced from their underlying drivers suggests something happened in the past, it doesn’t entitle you to make any sensible predictions about the future. There is a lovely old saying of statisticians (more recently picked up by computer modellers) that goes “Garbage in, garbage out”. I think it sums up the approach of Stockwell and Cox perfectly. Their poor understanding of climate science is only matched by the uselessness of their modelling. Perhaps I would be permitted to point to Figure 3 in the unpublished paper again. There are three straight lines on the graph, but only two are used to make any predictions. I have always wondered what was going through the minds of Stockwell and Cox when the third line was totally ignored.

I don’t think you have understood anything about Easterling and Wehner’s paper. They were asking an entirely different question: what is the probability that any decadal trend would have negative, zero or positive slope with or without global warming models. There analysis concluded there was a probability for decadal trends to be negative or zero even with global warming as modelled; however, such probabilities were much lower than in the absence of global warming. They then looked at the recent historical temperature data and showed the probability distribution was more similar to models including global warming. Nothing about break points. This is all about the climate variability being greater then the global warming signal over short periods.

Then you go on to link to a paper showing that climate variability is greater than the global warming signal over short periods! Well of course it is. But you can’t compare short periods with the century-long analysis of Swanson and Tsonis and draw any sort of meaningless conclusion.

Agronomist you say this:

"just because a statistical analysis of a set of numbers divorced from their underlying drivers suggests something happened in the past, it doesn’t entitle you to make any sensible predictions about the future."

How is that not what AGW presumes to do; make predictions on the basis of what has happened in the past, or is happening now? AGW, of course, is doubly fallacious, because it is based on fabrications of past and current data; so AGW predictions are based on current lies; the AGW predictions are therefore lies upon lies.

Anyway, you say this about Easterling:

"There analysis concluded there was a probability for decadal trends to be negative or zero even with global warming as modelled;"

Easterling uses a simple least squares model within decadal periods and assumes a symmetrical oscillation around a neutral mean with any overall trend above that attributable to AGW; if a decadal trend is negative, cooling, they assume a temporary superior contrary trend produced by natural variation within the decade.

That is nonsense for a number of reasons. Firstly it is well known that ENSO is asymmetrical:

http://farm3.static.flickr.com/2702/4503452885_79b5c09c4f_o.jpg

This produces a non-stationary element to ENSO which accumulates the conditions of the dominant ENSO phase:

http://www.cgd.ucar.edu/cas/adai/papers/MonahanDai_JC04.pdf

http://www.atmos.ucla.edu/~sun/doc/Sun_Yu_JCL_2009.pdf

In otherwords, ENSO, natural variability, produces trend.

Secondly, Easterling makes the same mistake that people who talk about 100 year floods make. Such predictions assume a constant probability of a flood over the 100 year period. We know that big floods are more likely during the -ve phase of ENSO, the -ve PDO, when La Nina dominates, and less likley during the +ve phase when El Nino dominates. If there are more +ve phases of ENSO, as there were during the 20thC, and they were asymmetrically bigger then of course trends would be greater as the 20thC progressed, simply due to that asymmetry in the natural variation.

Again this shows you do not understand the Stockwell paper which statistical test captures that asymmetry whereas Easterling does not.

cohenite, unlike the silly model of Stockwell and Cox, climate models use the known physical impacts of a variety of climate forces to drive the models. I can’t see how you can fail to see the difference, given how blindingly obvious it is. The predictions made from the model are based on known cyclical behaviours of the sun, known circulation of water and air, and known (and predicted) concentrations of CO2 in the atmosphere – among other factors.

I also see you have either failed to read Easterling and Wehner or failed to understand it. What the paper did was to take running 10 year periods through the recent climate record and through the predictions of several climate models and ask the question: what is the trend for that decade in degrees per year. From this they simply built a frequency distribution. You should read the legend to Figure 3.

No assumption of symmetric oscillation was made. This was a result of their analysis of the global models without greenhouse forcing. You do know the difference between an assumption and a result, don’t you?

As you have failed to understand what the paper was about, your other criticisms are moot.

I see you have trouble reading frequency distributions as well. If you want the one in 100 year event, you look at the extremes of the frequency distribution, not the average.

"The predictions made from the model are based on known cyclical behaviours of the sun, known circulation of water and air, and known (and predicted) concentrations of CO2 in the atmosphere – among other factors."

That's what Stockwell bases that paper's prediction on. I think it's amusing you have concentrated on the prediction graph from Stockwell and not the break concept and the strong statistical and climate evidence to support it; while at the same time extolling the virtues of AGW predictions which attempt to do the same thing with a less appropriate statistical model.

Anyway you say this about Easterling:

"No assumption of symmetric oscillation was made."

Easterling, page 5:

"Not surprisingly the probabilities for the long control runs are symmetrical around a zero trend, with more or less equal chances of a positive or negative decadal trend over the entire global surface air temperature time series."

On that assumption about natural variability Easterling then go on and declare that any trend must be due to AGW forcing; this conclusion is defective for the reasons I gave in my last post.

And you have the hide to say I haven't read or understood Easterling!

cohenite, you must have misread Stockwell and Cox. I suggest for punishment you should go and read it again. The model they proposed in Figure 3 is based entirely on 2 of the 3 lines they drew through the variable temperature data. It even says so in the legend to the figure.

“Prediction of global temperature to 2100, by projecting the trends of segments delineated by significant regime-shifts.” http://arxiv.org/abs/0907.1650v3

Nothing there about circulation models or anything else. The only stuff about underlying factors in the paper is post hoc. I am coming to the conclusion that you don’t understand statistical models at all, given the ludicrous claims you are making about Stockwell and Cox.

I see you have failed to understand Easterling and Wehner or are deliberately trying to mislead readers about the paper. I don’t really know how you manage this, the paper is after all only 3 pages long. http://www.agu.org/pubs/crossref/2009/2009GL037810.shtml

For the benefit of those reading, the bit you have cited from the paper is a conclusion from their analysis of the climate models without greenhouse forcing. It is not an assumption of their method. For you to continue to say so is wrong.

"the paper is after all only 3 pages long."

No, it is 13 pages long:

http://www.esrl.noaa.gov/psd/csi/images/GRL2009_ClimateWarming.pdf

About Stockwell you quote the predictive criteria used in that paper:

"Prediction of global temperature to 2100, by projecting the trends of segments delineated by significant regime-shifts.”

How have I been saying anything different?! It's a paper about breaks in climate conditions which the paper statistically connects with climate regime shifts and the attendant climate factors associated with such regime shifts; an attribute of climatology which is not controversial.

So, I'm wrong and can't understand statistics because the paper does what I say it did! That's just great!

And this:

"the bit you have cited from the paper is a conclusion from their analysis of the climate models without greenhouse forcing. It is not an assumption of their method. For you to continue to say so is wrong."

That's what I said!! The assumption of a zero trend around a symmetrical oscillation is how Easterling describes natural variation; it is their starting point with any trend above that zero assumption attributable to AGW. Their whole point is that in any period where there is no warming it is due to a period of natural variation which is compensating for the forced trend.

The foolishness of this idea is that in those periods where natural variation is the same as the trend the trend should be amplified to a trend which is the same as the trend is reduced in the periods when the natural variation is contrary to the trend. It isn't, so either the theory is wrong about the extent of AGW forcing or about natural variation or both.

In any event you are on another planet.