B. You could also do this study with a model for how you think such studies should be done (in terms of describing locations, methods etc.) Surely after all your criticism, you should have some opinions on how to do this right. (Might also be nice to see you cite some “model papers” that do type of job in recording things that you want in the esperimental literature.)

C. Regarding correspondance to local temp, I agree that you need to at least have foundational studies to show that there is even a proxy effect! And that such studies ought to give the relationship, linear or not, nonmonotinic, slope, etc. In general, I’m extremely disttrustful and coff at idea of “reflecting the global climate field” since the whole basis of the study is to build up local observations to a global picture. However, I do concede that their might be some “tele connections” via monsoon rain or what have you. But each individual one just needs to be proved in the foundational study (really no different in concept than the average tree ring calibration).

I note that you spend absolutely no time discussing the science or the plausibility of what Mann wrote (which makes no sense). I note that you do spend lots of time making ludicrous aspersions every time I show even a slight lack of respect for the work of people you clearly idolize. Mann, Bradley and Hughes are not gods, whose words are spoken from on high, but ordinary scientists promoted by a political process that is intolerant of criticism and apparently immune to the normal laws of logic.

I did not “describe M, B, and H as “nonsense’ and “magic’”, I note that Mann’s explanation of how trees can somehow not be sensitive to temperature but mysteriously can be sensitive to a "large scale of temperature" makes no scientific or rational sense whatsoever. It is not arrogant or dismissive to make such a statement. To say that Mann’s explanation makes no sense is not simply my view, but also Steve McIntyre’s (who referred to it as ‘magical’) and not a few scientists with degrees in the relevant disciplines.

It is typical of your hit and run tactics to continually snipe with these ludicrous charges and highly personalized attacks. Perhaps you should grow up a little and actually think about what is being presented.

Actually, make that "grow up a lot".

elsewhere we’re being urged to concentrate on the science not personalities, yet here you describe M, B, and H as ‘nonsense’ and ‘magic’. No reasons, no reasoning. Just a typically acerbic, arrogant dismissal of real and intelligent people. Noted….

The method only assumes that the signal in the predictor (not the predictor itself, which contains both signal and noise), varies linearly with some large-scale pattern of temperature, not with local temperature itself (see top right column of page 780 of MBH98, paragraph beginning with “Implicit in our approach”).

Does someone really have to have a PhD to know this is nonsense? The method is not reliant on science but upon magic. What on earth gives the idea that trees do not repond linearly to temperature but does to some “large scale pattern of temperature”?

Mann’s calibrations are to “temperature principal component” series rather than local temperature. So for example, for the Stahle Southwestern US/Mexico network, he uses 9 PC series in the AD1750 reconstruction, which he regresses against 11 temperature PCs. The regression period is only 79 years. So this regression is going to turn up a lot of high spurious correlations – as soon as you think about it, it is impossible to believe that there is a stable relationship between the PC7 of the Stahle/SWM network and the PC11 of world gridcell temperature, no matter what the correlations say.

By and large their statistical methods are very primitive. Mostly they rely on correlations, rather than checking t-statistics which are heteroskedasticity-autocorrelation consistent. In this area, economometric techniques are much more advanced than those used by the Hockey Team, so some of their condescension is pretty grating.

As I mentioned before, in our Nature correspondence, Mann argued that MBH98 did not assume that proxies have a linear relationship to temperature as follows:

“To make matters worse, they attempt to do so based on an incorrect description of the assumptions behind the MBH98 methodology. They claim that the method requires that predictors have “a linear relationship to temperature”. The method only assumes that the signal in the predictor (not the predictor itself, which contains both signal and noise), varies linearly with some large-scale pattern of temperature, not with local temperature itself (see top right column of page 780 of MBH98, paragraph
beginning with “Implicit in our approach”). For example, a coral indicator in the western tropical Pacific which records precipitation influences due to the El Niño/Southern Oscillation is a suitable proxy for ENSO-related sea surface temperature patterns. This issue is discussed both in MBH98 and numerous follow-up articles by the authors. The demonstration by MBH98 of Gaussian calibration residuals indicates that the linearity assumption of the MBH98 reconstruction is not violated.”

So there’s not much statistical analysis. I’ve been trying to get the residuals for the 15th century construction step for over 18 months without any success. The “demonstration” method is remarkably simplistic.

Regards, Steve

You say that there is no attempt by many of the hockey team to relate growth changes to local temperature but just to some theoretic NH climate. If that is true (I would be amazed) how do they calibrate the tree rings? They must have some way of calibrating each species response to temperature (using the local instrument record?) otherwise all you can say is year X was warmer than year Y because the rings are wider. Indeed you may not even be able to say that because I dobt the response is linear.