I’m more inclined to place the differences on the properties of the proxies themselves. Until you do runs with pseudoproxies that more realistically represent MBH proxies – contaminated bristlecones, some proxies correlated to precipitation, some just noise period with no signal, then you haven’t really modeled the MBH dog’s breakfast. I think that that’s where a reconciliation will come.

FWIW, I have to second this. When you actually look at the raw proxy data (which Steve has been very good about posting) you see how bizare the underlying data actually is. Using well-behaved time series to model them doesn’t begin to capture their diversity. Perhaps it would be more realistic to generate the pseudo-proxies using a number of different engines with very different properties.

I must confess to feeling a certain amusement at Mann savaging VZGT for allegedly “incorrectly” implementing his precious methodology. Back in 2003, when we sought clarification of MBH methodology, Mann refused, on the basis that von Storch and Zorita had found his existing disclosure sufficient to implement his methodology (see Mann correspondence). In the Corrigendum SI, Zorita et al 2003 is cited on 2 different occasions as an accurate implementation of MBH methodology.

Yes, quite laughable how some can speak out of both sides (whichever suits them at the time!).

It’s pretty amazing how they complain on the one hand about people criticizing a “10 year old paper” – with Mann, arithmetic within 25% is pretty good – and then continually pick at scabs by publishing stuff like Wahl and Ammann [Climatic Change] and Wahl et al [Science].

Yep, amazing and amusing…wonders will never cease!

That paper is typical. Figure 1, supposedly showing modelled versus actual temperatuere data make it a challenge to judge the quality of the output, beyond “eyeballing2 (TM Hockey team).

I wouldn’t have been too much trouble to use instead a similar figue showing the deviations as “modelled-actual”. That would probably be a bit emarassing though.

Epica, that a model shows increased temperatures in the 20th century is a necessary, but not sufficient outcome of a
model. As far as the instrumental record is reliable, the model outcome must follow the same pattern, or should be discarded. That doesn’t mean that models have any predictive power, as there are too many sets of parameters which all can give the same “curve fitting” of the past century/centuries, from 3-5 times direct solar (at the cost of the CO2/sulfate tandem) to positive and negative responses of cloud cover to increased temperatures… After all we have only one equation with one/two dependent (temperature, precipitation) with many independent (solar, volcanic, GHGs, aerosols,…) variables and with interdependent feedbacks like cloud cover, air/ocean circulations)…

That even state-of-the-art models have troubles to reflect reality, can be viewed in figure 1 of Johanessen ea. where the ECHAM4 model underestimates the measured temperatures in the 1930-1940 period, and overestimates the influence of GHGs and aerosols in the second period (and thus overall for GHGs)…

I’ve never used words like "communist greenhouse conspiracy" or anything remotely like that. I do not do yell "frawd" "all the time" and have intentionally stayed away from any such suggestions. I have repeatedly said that, for policy purposes, one should rely on a consensus; however, for scientific purposes, every detail should be probed. In terms of data analysis, I think that the record of this blog is exemplary.

I do not believe, for example, that the articles, Mann and Jones 2003; Jones et al 1998; Briffa et al 2001 and MBH, for example, are "independent" in terms that a civilian would understand. Your outrage should be directed at people making the ludicrous claims that these studies are "independent". Being non-independent doesn’t mean that the studies are "wrong". Each one stands or falls on its own merits. But don’t tell me that they are "independent".

Yes, I would have liked a citation. Eduardo is a very decent guy and I’m sure that he’s a little embarrassed about the non-citation, but I’m not going to get into a war with him about it. He, Fidel and Hans have been nice to me privately and, while I would appreciate more public support e.g. a citation here, they’ve been virtually alone in the climate science community in having spoken favorably about me in the past.

So Jean, while I would have liked a trick here, I can live without it. There are more cards to be played. I appreciate the support, but I’m confident that things will sort out under more favorable conditions.

Eduardo, if you’re reading this, I’m working on a long and interesting note synthesizing these issues, which I think that you’ll be interested in. If you’re online, why don’t you wait for that and we won’t pick at scabs.

Now, since you are here, would you mind answering the following question.
In your Science (von Storch et al.) paper (combined with the response), your key contribution, in my opinion at least, is that you show, using simulated temperatures and proxies, that the “regression methods” used in MBH are flawed since they underestimate the pre-caliberation period. Moreover, it is evident from your response, that this is especially true if the noise is colored. These are, in my opinion, also the key points raised by Steve. So my question, as a working scientist, to you is that how do you justify the fact that you do not cite Steve’s results?