“Ensemble Of Climate Models Is Statistically Meaningless”

A new posting on WUWT titled  “The “ensemble” of models is completely meaningless, statistically” by Robert G. Brown, Duke University Physics Department is getting a lot of favorable  attention.  He says that if you know statistics, you would recognize that the 73 CMIP models grouped together has no meaning.(See my posting here for more info on CMIP models.)

I am not sufficiently versed in statistics to make an assessment of his work, but I believe the sense of his comments is that the climate models’ outputs are really not actual data so you can’t do a statistical analysis as if they were.  To illustrate his contention that the warmers are using a “mean” derived from an ensemble of climate models to make a  forecast of future temperature, see this chart. And he concludes that something like 68 of the 73 CMIPs should be retired.

Brown says:

What I’m trying to say is that the variance and mean of the “ensemble” of models is completely meaningless, statistically because the inputs do not possess the most basic properties required for a meaningful interpretation. They are not independent, their differences are not based on a random distribution of errors, there is no reason whatsoever to believe that the errors or differences are unbiased (given that the only way humans can generate unbiased anything is through the use of e.g. dice or other objectively random instruments).

So why buy into this nonsense by doing linear fits to a function — global temperature — that has never in its entire history been linear, although of course it has always been approximately smooth so one can always do a Taylor series expansion in some sufficiently small interval and get a linear term that — by the nature of Taylor series fits to nonlinear functions — is guaranteed to fail if extrapolated as higher order nonlinear terms kick in and ultimately dominate? Why even pay lip service to the notion that or for a linear fit, or for a Kolmogorov-Smirnov comparison of the real temperature record and the extrapolated model prediction, has some meaning? It has none.

Let me repeat this. It has no meaning! It is indefensible within the theory and practice of statistical analysis. You might as well use a ouija board as the basis of claims about the future climate history as the ensemble average of different computational physical models that do not differ by truly random variations and are subject to all sorts of omitted variable, selected variable, implementation, and initialization bias. The board might give you the right answer, might not, but good luck justifying the answer it gives on some sort of rational basis.

In my posting, previously mentioned, I said:

It is very likely that most, if not all of the modelers, are very intelligent people.  When I see the mess of curves that these modelers produce, one thing always occurs to me—If you could actually model the atmosphere to get a forecast temperature, you would only need one model.

Any more that than shows that all these bright people have 73 different ideas of how the atmosphere works and not one of them can duplicate actual measurements.

cbdakota