Jones and CRU exonerated

[Poptech]

If you define "in the tank" as roughly 57% of people accepting the climate science consensus, sure. I know that's down quite a bit from the high somewhere in the 70s, but it still isn't awful (at least those are the numbers the last time I looked). Look at similar polls in the middle of summer and I'm certain we'll see an uptick.

Tanking Matt, better try harder!

Americans' Global Warming Concerns Continue to Drop

Push for the whitewash reports faster, faster!!! I am not sure if you can save it in time.

 

[uke2se]

Tanking Matt, better try harder!

Americans' Global Warming Concerns Continue to Drop

Push for the whitewash reports faster, faster!!! I am not sure if you can save it in time.

And, once more, Poptech provides ammunition for any future argument about his stance on science.

 

[MattusMaximus]

Tanking Matt, better try harder!

Americans' Global Warming Concerns Continue to Drop

Push for the whitewash reports faster, faster!!! I am not sure if you can save it in time.

Ah yes, "science by polling" courtesy of PopTech.

Well, using that as a guide, I suppose here in the U.S. we really should scrub evolution and the big bang out of all the textbooks.

 

[Poptech]

I suppose here in the U.S. we really should scrub evolution and the big bang out of all the textbooks.

How does straw taste strawman?

 

[uke2se]

How does straw taste strawman?

Surprisingly little like straw.

Shouldn't come as a surprise as it isn't a straw man fallacy.

 

[Poptech]

Shouldn't come as a surprise as it isn't a straw man fallacy.

So far you have demonstrated to be the most intellectually inept person trying to argue a point on these forums.

 

[MattusMaximus]

How does straw taste strawman?

Interestingly, a lot like most of your arguments

 

[uke2se]

So far you have demonstrated to be the most intellectually inept person trying to argue a point on these forums.

If I didn't know any better I could swear you're trying to hurt my feelings.

 

[CoolSceptic]

Alright, this is silly.

Yes, your entire argument is silly, and a pointless straw man. I'll explain why here.

Either you're being intentionally dense or you've so totally missed the point of Halley's paper that there's no reason to continue.

No, I've not missed a thing about Halley's paper. My original claim in this thread was *NOT* that AGW doesn't exist. I even pointed this out explicitly in reply to a question long before Halley was ever raised.

My original point was that LTP exists in climate series, and must be accounted for in order to determine whether effects are significant or not.

As noted in Cohn and Lins, a previously assumed given (that the trend in the instrumental temperature with a p-value of 10^-7) becomes actually uncertain, to the point we have to use an entirely different method, and the significance becomes marginal.

This doesn't just apply to this one result though. LTP needs to be accounted for in all climate time series analysis. And it just isn't. Which means things that were once assumed highly significant may not be any more.

On this point, Halley's paper is fully in agreement with me. As is Cohn and Lins. As is Rybski et al. As is Koutsoyiannis. As is Mandelbrot, Klemes and Hurst.

All these papers AGREE with me on this point. Not one paper has been raised that REFUTES this position.

Shoot him an e-mail. Ask him what he thinks about the "bias."

LOL. No. Because you're the one obsessing over the bias. To me, it is a side show straw man that misses my point entirely.

Ask Halley if he agrees with my statement above. That's the interesting question which has serious implications for climate science. Unfortunately, because you've gone down an irrelevant rabbit hole, you can't see the wood for the trees.

 

[CoolSceptic]

Again, from Halley:

<same passage already quoted over and over snipped>

The magnitude is unkown

Right! THE MAGNITUDE IS UNKNOWN.

As I've pointed out over and over again.

In science, you don't get to say "this is unknown, but it must be small". If it is unknown, it is unknown. Full stop.

Your way is not how science works.

I appreciate you don't get this...

but it's unlikely that the warming trend can be explained with natural variance. That's all I've said.

And that statement is just plain wrong. The magnitude the variance must change is LESS THAN THE EXISTING SPREAD BETWEEN RECONSTRUCTIONS. That isn't "unlikely" by any stretch of the imagination. It is very plausible. So plausible, as I've already demonstrated, IT HAS ALREADY HAPPENED ONCE BEFORE.

I've already pointed this out. I've shown you the numbers. Yet you ignore this point over and over again, put the blinkers up and just keep insisting that it is unlikely, even though IT HAS ALREADY HAPPENED ONCE BEFORE.

Just answer that point. You won't, because you can't.

That's what Halley says, despite your comical attempts to pretend like he argues something else.

Wow, your ignorance is staggering. Halley argues a lot of different things in that paper. I've already stated that on just one point - the word unlikely - I disagree with him, so I'm clearly NOT saying Halley is arguing something else.

On all the other (much more interesting points) he makes in the paper, I am in full and complete agreement with Halley. But you create a ludicrous false dichotomy that Halley's entire paper boils down to whether the bias in reconstructions is such that we can claim 4.99 or 5.01% p-values. That is the result that is not evenly remotely interesting. The interesting result - as I've said over and over again, which Halley, Rybski and everyone else fully agrees with me on, is the presence of LTP and the effect of it on confidence intervals.

Yet you can't even see this conclusion exists, because you can't see the wood for the trees.

Time will tell whether your buddies can gin up more significance concerning their statistical theory.

Time will tell whether you will learn how to read scientific papers one day. It isn't obvious to me that you do understand science, but this is a classic:

Exact magnitude unknown, but I bet I'm right.

This one sentence just says it all. We don't know, but you know you are right. Yep. That's how science works.

 

[CoolSceptic]

Halley calculated how much larger the variability would have to be in order to prevent the natural-variance-explains-the-instrumental-trend hypothesis from being rejected at a p-value of 0.05.

Yes. I know. I've already explained this.

That gives us a quantitative lower bound for the size of any postulated bias that would be worthy of consideration.

Yes, and there is a reasonable case to argue that the correct value is a variance of 2.7x, not 4x, because the difference between TaveNH2v and CRUTEM3 is really a form of uncertainty. Halley does make this point in his paper (but unfortunately repeats the incorrect "4x"). You should also note this is a variance, not a standard deviation, so has units of deg C². So the scale error in reconstruction units would be x1.65 and x2.

As things stand, with no good evidence for a bias in the dArrigo2 series

There is good evidence to assume this. D'Arrigo use variance matching, which causes a known bias, as outlined in von Storch et al. These papers show a scale error of x1.65 is indeed well within possibility, and in fact a scale error of x2 (50% reduction of spread in the reconstruction) is quite likely.

Also, we have an indication of the magnitude of uncertainty by looking at the spread of inconsistencies in the reconstruction, which also exceeds the values being discussed. So, in fact, we already have evidence of problems of this magnitude. As I said, this would give most scientists pause.

Halley concludes that the probability of the observed warming trend being generated by natural variability is low or very low (p < 0.01 for TaveNH2v, p < 0.001 for CRUTEM3).

But needs to make assumptions about the bias to conclude this, which cannot be quantified. So the conclusions make assumptions which are trivially shown to require further research. This is no different from the conclusions of Koutsoyiannis and Montanari 2007.

The important thing to note is the difference between conventional climate science methods, which argue a significance of 10^-7, and the approach adopted by Halley, which is much more uncertain. Whilst the question of whether AGW exists or not is a less interesting one, this main observation has serious consequences across climate science.

Halley makes this clear in his paper. I appreciate most here, as I've already said, will not be able to see the wood for the trees.

 

[macdoc]

will not be able to see the wood for the trees.

the irony, the irony..._{.with apologies to Brando}

 

[TraneWreck]

No, I've not missed a thing about Halley's paper. My original claim in this thread was *NOT* that AGW doesn't exist. I even pointed this out explicitly in reply to a question long before Halley was ever raised.

My original point was that LTP exists in climate series, and must be accounted for in order to determine whether effects are significant or not.

As noted in Cohn and Lins, a previously assumed given (that the trend in the instrumental temperature with a p-value of 10^-7) becomes actually uncertain, to the point we have to use an entirely different method, and the significance becomes marginal.

This doesn't just apply to this one result though. LTP needs to be accounted for in all climate time series analysis. And it just isn't. Which means things that were once assumed highly significant may not be any more.

On this point, Halley's paper is fully in agreement with me. As is Cohn and Lins. As is Rybski et al. As is Koutsoyiannis. As is Mandelbrot, Klemes and Hurst.

All these papers AGREE with me on this point. Not one paper has been raised that REFUTES this position.

Ah, this is cake + eating.

Halley FLATLY says that even when LTP is considered natural variance cannot explain current warming on ANY model. Because the wide range of models are all in agreement on this point, the only remaining explanation is one of bias, a point he ridicules.

His references for people arguing bias are to 1) Congressional Testimony where a scientist makes an ass out of himself, not peer reviewed work. I gave you a link that explained that, and 2) to another controversial author who has resigned from a journal over the publishing of a terrible article and had his work severely criticized. I then gave you two peer reviewed articles that bring his methodology into question.

I know subtlety is not your forte, but if you fail to understand the degree to which Halley holds the idea of bias in total contempt, I don't know how to help you.

Halley and Rybski find natural variance, even when LTP is considered, to be insignificant to explain current warming.

You introduced this as part of your criticism of the IPCC. So to sum up: according to you, the IPCC, like the Iraq dossier, cannot be trusted because 1) a minor typo corrected later in the paper, 2) an unpublished graph that was used correctly, and 3) because they didn't consider the work of Cohn and Lins seriously enough even though the result of involving LTP in climate models doesn't undermine AGW.

Wow, that was worth the effort.

 

[TraneWreck]

This one sentence just says it all. We don't know, but you know you are right. Yep. That's how science works.

Haha, wow. Just wow. So you think it's possible that Halley is 8'5"?

Yes, we may be unsure of magnitudes within a certain degree of error, and yet reasonably sure on other scales. This is Halley's point. I'm not going to even elaborate because you so continually miss this simple conclusion.

And remember, the 4x number is only relevant if you assume that the most favorable of the models is true. If you take an average, then the amount of variance for natural changes is farther outside of that number than the models internally disagree.

Halley explains why he is unmoved by the point you keep wailing about.

 

[W.D.Clinger]

Yes, and there is a reasonable case to argue that the correct value is a variance of 2.7x, not 4x, because the difference between TaveNH2v and CRUTEM3 is really a form of uncertainty.

Part of the difference is a form of information: Halley's CRUTEM3 data extend several years past TaveNH2v.

As things stand, with no good evidence for a bias in the dArrigo2 series

There is good evidence to assume this. D'Arrigo use variance matching, which causes a known bias, as outlined in von Storch et al. These papers show a scale error of x1.65 is indeed well within possibility, and in fact a scale error of x2 (50% reduction of spread in the reconstruction) is quite likely.

If you're referring to von Storch et al 2005, then their "may be" and "can result" have become your "quite likely":

The estimate of this variance may be also subject to uncertainties. It has been previously found (Esper et al. 2005) that, in the observational record, the choice of calibration period to estimate the observed variance of the target variable can result in differences in the amplitude of the final reconstruction of up to 50% of the total amplitude.

Furthermore von Storch et al's reference to the "choice of calibration period" appears to be a reference to paragraph 27 of Esper, Wilson, and Briffa 2005, which says nothing about a 50% difference. The difference of up to 50% arises not from the choice of calibration period but from the use of regression instead of scaling, as discussed in Esper et al paragraphs 11-14 and summarized in paragraph 24. Wilson was a co-author of d'Arrigo, Wilson, and Jacoby 2006, which cites Esper, Wilson, and Briffa 2005 to explain why they use scaling (see paragraphs 10 and 11).

In short, I don't see how von Storch et al can be used to argue for a known bias in d'Arrigo et al, let alone that a factor of 2 is "quite likely".

The important thing to note is the difference between conventional climate science methods, which argue a significance of 10^-7, and the approach adopted by Halley, which is much more uncertain. Whilst the question of whether AGW exists or not is a less interesting one, this main observation has serious consequences across climate science.

Halley makes this clear in his paper. I appreciate most here, as I've already said, will not be able to see the wood for the trees.

Yes, Halley's p-values are more realistic.

Yesterday's Report of the International Panel set up by the University of East Anglia to examine the research of the Climatic Research Unit contains four numbered conclusions. The first exonerated CRU of deliberate scientific malpractice. Here is the second:

We cannot help remarking that it is very surprising that research in an area that depends so heavily on statistical methods has not been carried out in close collaboration with professional statisticians. Indeed there would be mutual benefit if there were closer collaboration and interaction between CRU and a much wider scientific group outside the relatively small international circle of temperature specialists.