Re: Jeff Alberts (#801),

Of course Jeff . What I post is public and I claim no copyrights ðŸ™‚

]]>Re: Craig Loehle (#800),

Craig

“If you try to explain away the lack of warming the past 10 yrs with the endogneous fluctuations model, you are treading on thin ice because this implies that your attribution of warming in the 1980s-1990s to GHG might have been spurious. You can’t have it both ways.”

Thank you. I had asked that question about 10 days ago, ie. if current decadal NON-warming is due to oceanic latency (endogenous fluctuation), why is the the previous 20 years of warming not exhibiting the same latency ?

GCM’s cannot have it both ways, as you say.

]]>Re: bender (#805),

In my admittedly extremely uninformed opinion, of you don’t know and understand all of the parameters 100%, how can you know you’ve reached the “right” conclusion for the “right” reason? I mean, the model might end up with a global temperature (as meaningless as that is) that “matches” observations, but there’s no way to know if the model is “right”. It’s just another correlation, and this time, totally man-made.

]]>Re: Dan Hughes (#804),

it seems to me that to insist that your favorite GCM is correctly exhibiting chaotic response and at the same time insisting that a trend exists is contradictory

That, I do not necessarily agree with. And this is not a “correction”, but personal conjecture – which Steve M has repeatedly stated he is not interested in. I think the two – chaos + external forcing trend – can co-exist in a data series, that it is not “contradictory” to suppose this is the case, eiither in a model or in the real climate system. IMV the problem is quantitative: efficient separation/estimation of the two independent components. In a model it is a no-brainer because you specify the deterministic forcing trend and you can study the deterministic chaotic noise in the absence of external forcing. In reality, well, that is the challenge. You do not know that your model is correct. In fact you’re pretty certain it’s not, because of the aerosols uncertainty, if nothing else.

.

IOW although I think attribution is theoretically possible, I don’t think it’s as easy as GS makes it out to be. But this comment of Hansen’s about choas cascading across all time scales – it is very intriguing. One wonders what Tom Vonk and Dan Hughes make of the papers cited by Hansen in support of this view. (See #381 above.)

I have yet to get a firm handle on the weather vs. climate when chaos is the subject. My understanding is that weather is chaotic and climate is not, but any given trajectory from a GCM calculation is. I assume the latter refers to, for example, the temperature. Maybe it refers to all dependent variables?

From time-to-time in his comments, Tom has mentioned that chaotic response cannot exhibit a trend. A trajectory of a dependent variable displayed as a function of time cannot have a trend with respect to time. The plot will show an aperiodic, bounded, line that neither grows or decays with time. Again this is for the case of constant parameters; if the parameters vary, as the corresponding quantities in fact do in the real-world, all bets are off. The aperiodic nature of the plot can if fact cease. It is also worth repeating that for all the classical systems of ODEs for which chaotic response is known, that response is obtained for only subsets of parameter space.

If a trajectory from a GCM calculation is presumed to represent chaotic response, how can that trajectory display a trend? Maybe chaotic response when spatial-temporal systems are the subject? I don’t know. But it seems to me that to insist that your favorite GCM is correctly exhibiting chaotic response and at the same time insisting that a trend exists is contradictory.

Chaotic response always requires that energy be continuously added into the system; otherwise the trajectory will approach a null equilibrium state. When the weather is the subject of discussion, generally, in the real world, energy is not continuously added into the system at specific locations. Certain conditions act to remove energy from the atmosphere. It happens almost every night, for one example.

Generally, the extrapolation of what is known about chaotic response from small systems of simple non-linear ODEs, to the case of GCMs, and further to the real-world based solely on GCM (or NWP ) results, is a stretch that cannot be supported by a theoretical foundation.

And how ODE chaotic response is further extrapolated to the climate, which is a system of inherently complex interacting subsystems, is in my opinion, the purest of hand-and-arm waving. As Tom has mentioned, averaging of chaotic response is not on any firm foundation. And being that climate is defined as a long-term average of the weather, I suspect we have really jumped into the deep end of the pool.

All corrections will be appreciated.

]]>Re: Craig Loehle (#800),

I’ve said it several times before but it is worth emphasizing: 20th century warming came in two pulses and the first of these in the 1930s was marked by extremely warm Arctic anomalies that currently have **no explanation **– by *Hansen’s own admission*. Not GHGs, not BC (black carbon). If this is “internal variability” at work then I assert that the Team consistently underestimates this component. If this component is large and unpredictable then I couldn’t agree more with Dr. Loehle: attribution would not be a slam dunk. The statistics would be pretty dodgy. The continual twiddling of that aerosol knob – and the denial over its role in GHG attribution – is highly disconcerting given this dodginess.

Re: TomVonk (#799),

I’m happy to report that this does not conflict with my superficial understanding/assumptions about how chaos theory applies to climatology. I sure wish Hansen & Schmidt et al. would engage on the topic of statistical climatology.

Re: TomVonk (#799),

Tom, do you mind if I post this on my web site http://whatcatastrophe.com ? you are also welcome to register there and post it yourself, if you’d like to change the formatting or anything.

thanks in advance,

Jeff

]]>