I'm dead. Send flowers

Congratulations to David Hoffer for his idea. We did something similar when I was at Churchill College. The College, which liked to appear trendy, erected various bits of distorted tin all ove the College. So we got our mates in the Engineering faculty to run up half a dozen similar clusters of scrap left over from a hyperdrive motor they were building, which were erected all over the College one night.
The next morning I asked the Bursar whether he could tell the difference between the real ones and the fakes. He admitted he couldn’t. That’s because they’re all fakes, I said. He got the point. They were gone by later that day. He sold the whole lot for scrap. Got quite a good price for the dilithium, by all accounts.

Lord Monckton has worked to give the impression that there was something underhanded in my observation that Monckton et al. haven’t been forthcoming about precisely how they derived their transience-fraction values from Roe. Since he has publicly called my intellectual honesty into question, permit me some personal observations about that.
Several commenters had raised questions about the transience fraction on various threads in which Lord Monckton had participated and given (largely, non-) answers. I therefore wrote a post directed to the questions raised by that quantity and stating that the manner in which transience fraction had been derived from the Roe paper wasn’t entirely clear. In the ensuing thread Lord Monckton made a merely conclusory comment that transience fraction is unity for , and I responded that he could test that proposition’s plausibility by plotting the implied step responses ostensibly obtained from Roe. Lord Monkcton’s subsequent post stiff-armed my post’s question with the (non-) answer that “The table was derived from a graph in Gerard Roe’s magisterial paper of 2009 on feedbacks and the climate.” I then posted R code that would actually provide a plot that illustrates how those values seem inconsistent with each other. I subsequently stated yet again that “Monckton et al. don’t tell how they got from the plot to their table.”
After all this occurred in threads in which Lord Monckton was actively participating and responding (however inappositely) to comments, I submitted a comment expressing my disappointment, in light of the Table 2 rows’ apparent inconsistency with each other, that Lord Monckton had insisted on withholding the requested information about precisely how Monckton et al. inferred the Table 2 values from the Gerard Roe paper. I then submitted a further post, in which by linking “turned down” and “request” to his and my posts, respectively, I characterized his post as turning down my post’s request for that information. (At least one of Lord Monckton’s apologists later actually professed to understand from that link of “request” to my previous post that by “request” I meant an e-mail message, and Lord Monckton jumped right on the theory that an e-mail message is what I had falsely implied. This from someone who presumes to lecture me about intellectual honesty.)
In response, the authors then objected in a further post that “Mr. Born at no time contacted any of us to ask for the information he now says we are ‘withholding” and “refusing to provide.’” Of course, that was disingenuous. In view of the back-and-forth on the various threads in which he actively participated, Lord Monckton could not possibly have been unaware not only that I had asked, repeatedly, for more information but also that I had identified discrepancies that had motivated me to do so. As the repetition of Lord Monckton’s non-answer in that post suggests, moreover, repeating the request through some alternate channel would almost certainly have proved futile; as they did in that post, they would simply have continued their stonewalling.
Let’s look at what they were withholding.
According to their paper, the authors’ intention was that their “model” be used to “illuminate” more-complex climate models. Since the IPCC did not provide its models’ step responses (“evolutionary curves” in Monckton et al.’s parlance), Monckton et al. purportedly “derived” proxies from the Roe paper. They normalized Roe’s step responses and placed the resultant transience-fraction values in their Table 2. Models defined by the thereby-specified step responses are what they plugged into their Eq’n (1) to get all their example results: if the Roe responses are representative, they obtained outputs something like those that the more-complex models would yield under the same feedback assumptions. So it was important to know how they arrived at the transience-fraction values, especially because the relationships among them were puzzling.
Their paper says explicitly that each value in each of Table 2’s five rows came from the Roe paper, but it doesn’t say how. Specifically, Table 2’s heading says: “Approximate values of at values and = 0.5, 1.3, 2.1, and 2.9 over periods t = 25–300 years, derived from [the Roe paper]”: each period’s value in each row is “derived from” Roe. But “derived from” is very much open to interpretation, particularly in view of the fact that there are at most three curves, not five, from which to eyeball values in the Fig. 4 graph, by which Monckton et al. illustrate Roe’s step-response range. Did something in Roe’s text provide a basis for the other two rows? I wasn’t able to identify anything. So, again, “derived from” doesn’t tell precisely how they got the Table 2 values. That’s the information I sought.
Now, Lord Monckton’s reply post did seem to add a little information. “The table was derived from a graph in Gerard Roe’s magisterial paper of 2009 on feedbacks and the climate,” he said. “Far from obscuring anything, we had made everything explicit.” Well, okay, then; everything’s explicit: nothing came from the text; all five rows somehow came from that graph.
This tells us that “derived” has to mean something more than just eyeballed; as we noted previously, there are more table rows than there are graph curves. We could speculate about how the “derivation” was carried out. In my second post, for example, I laid out how I had generated five curves from one by making certain assumptions, and four of the resultant curves approximated the values in four of Table 2’s five rows. But the remaining curve didn’t match the first row’s values, so Monckton et al. must have made assumptions different from mine. Why couldn’t they just tell us what they were, as I had in the case of my curves? Why do we have to speculate? I felt like Steve McIntyre trying to get information from Michael Mann.
Moreover, the assumptions they would have had to make are hard to imagine. What set of assumptions would imply that more feedback ever results in less response, as the first three entries in Table 2’s first column indicate (after multiplication by static closed-loop gain to denormalize them)? Actually, I can imagine such assumptions, but they are quite contrived. So those entries at least seem implausible, as I had already pointed out three times when the authors replied in their second post.
And what was the reply?

In our paper, Table 2 gives approximate values of the transience fraction corresponding to equilibrium feedback sums f ≤ 0 and f = 0.5, 1.3, 2.1 and 2.9. Where the equilibrium feedback sum is less than or equal to about 0.3, the transience fraction may be safely taken as unity: at sufficiently small f there is little difference between instantaneous and equilibrium response. For f on 2.1 [1.3, 2.9], the value of the transience fraction is simply the fraction of equilibrium sensitivity attained in a given year after the initial forcing, as shown in Roe’s graph, reproduced at fig. 4 of our paper.

That is, they merely repeated a portion of what their paper had already said. Is that the reply of authors who are honestly trying to explain precisely how they got five curves from at most three? I don’t think so.
Indeed, if it weren’t for their Table 2 heading’s explicit statement that all the f ≤ 0 values were derived from Roe, the “Where the equilibrium feedback sum is less than or equal to about 0.3, the transience fraction may be safely taken as unity” passage would seem intentionally ambiguous. To be consistent with Table 2’s heading, it has to be interpreted as an observation about the values derived from Roe, but one might otherwise have interpreted the reply as saying that the f = 1.3, 2.1 and 2.9 values were eyeballed from the graph but that the f ≤ 0 values were merely the authors’ assumption, that they had just made them up.
Now, I personally think there’s some evidence that at the coarse time resolution Monckton et al. used in their paper it’s a reasonable approximation to treat the climate system as memoryless (as far as the forcings of interest are concerned). But basing Table 2 entries on that assumption would have been a product of muddled thinking. To “illuminate” the “more-complex models,” Monckton et al. were ostensibly approximating them with Roe-like models, so the Table 2 entries needed to come from Roe, as the heading said they did; they couldn’t be just the authors’ postulate.
Yet I now suspect that’s exactly what they are, that Monckton et al.’s Table 2 heading is a falsehood. A subsequent intemperate comment (It was full of puerile language, misrepresentations of my position, and attacks on me, my honesty, and my punctuation) contained the following statement:

He wails that Roe’s paper has only one (or maybe three) curves but our table 2 had not one or three but five sets of values. It is plain to anyone looking at Roe’s graph that there are three curves on it. And it is also plain – and explained in the text of our paper – that our fourth set of values – unity where the feedback sum is sufficiently low – is an approximation that will not lead to significant error. As to the fifth set of values, it falls between Roe’s least curve and the very-low-feedback case.

Obviously, this is as Delphic a statement as his others about those values’ provenance; like them it could indeed be interpreted in a manner consistent with Table 2’s heading: it could be read merely as observations about value the authors obtained in some way from Roe. But to me it sounds like an attempt to provide a basis for later contending that they had all along said they’d eyeballed the graph to get the last three rows, postulated the first the first row’s unity values, and in an undisclosed manner interpolated between the first and third rows to get the second. (Comparison of the second row’s values with the first’s and third’s reveals that the interpolation approach was, well, whimsical.)
If recent history is any guide, Lord Monckton will probably dismiss the Table 2 values’ provenance as a “quibble.” I don’t think it is. Nearly all their paper’s conclusions are presented as the results of a new model that was developed over eight years of effort. Those included their featured Fig. 6’s alleged demonstration of skill. By not unambiguously identifying the first-row entries’ provenance they tended to obscure how little the §8.3 calculation on which Fig. 6 was based follows from the logorrhea that preceded it.
Without the transience fraction’s derivation from Roe, the only contributions the rest of the paper makes to that calculation (which §8.3.5 summarizes) are RCP 6.0 values of forcing and the IPCC’s estimate of open-loop gain, or “Plank climate-sensitivity parameter ”—both of which came from the IPCC, of course, not from the authors. The authors contributed only a feedback range and the transience fraction from Table 2’s first row. Despite a nod in the direction of their Fig. 5, that feedback range’s lower limit appears to have been completely arbitrary, and its upper limit came from the almost equally arbitrary “process engineers’ design limit.” If my suspicion is correct, moreover, the transience fraction is merely a value they postulated.
So their featured Fig. 6 represents only the unremarkable conclusion, dictated merely by the definition of feedback, that the RCP 6.0 forcing level implies a temperature change that’s quite modest if the “process engineers’ design limit” and the IPCC’s open-loop-gain estimate are assumed. That recognition didn’t require a new, “irreducibly simple climate model”: the elephant had labored and brought forth a mouse.
Now, the assumption of some lower transience fraction instead of the one they postulated would not have detracted from the Fig. 6 conclusion. But that’s not the point. The point is that their transience fraction is a key quantity in their calculations, that I asked from the beginning how, exactly, Monckton et al. “derived” its values from Roe, and that they have persisted in withholding that information. Lord Monckton keeps misrepresenting my position as being that I want a more-complicated “model” or different transience-fraction values. It should be apparent by now that instead I wanted an explanation that’s clear and coherent. Such an explanation should have been easy to give.
When I accordingly remarked on their conduct the authors reacted as follows:
To allege that authors of a scientific paper have deliberately withheld requested information is to make a very serious allegation of professional misconduct. That is the allegation that Mr Born has now made twice, and in the bluntest terms. . . . He must withdraw that allegation, and be very careful in future not to repeat it.
Well, only Monckton et al. know whether their conduct was deliberate, but their withholding that information is not just an allegation; as we saw above, it’s a matter of record. To withhold means to refrain from giving. Did any of their content-free responses give the requested information? No; merely repeating where the values came from but not telling how is not giving the information. Giving the information would have been saying something like, No, our table heading is false; we really eyeballed only the last three rows from Roe; the rest we made up. They didn’t do that: they withheld the requested information. I leave to others the question of whether that constitutes, as Monckton et al. put it, “professional misconduct.”

Methinks Mr Born doth protest too much, and off topic, and to no avail. Off topic, because the reference to the paper at scibull.com in the head posting was there solely to establish that in that paper we implicitly acknowledge throughout that there is a greenhouse effect and that Man has some influence on climate; and to no avail, because he said we had “refused” to supply information he had not contacted us to ask for. That was intellectually dishonest, as several commenters here pointed out: and Mr Born’s criticisms of our paper came across as “mean-spirited”, as another commetner pointed out.
The transience function is a simple enough concept. Nothing central to the paper turned on it. Our values are self-evidently a fair representation both at instantaneity and at equilibrium, and tend somewhat to overstate climate sensitivity conservatively at all points in between, and the reader is specifically told he is free to choose his own values, and that is all that needs to be said. I could argue in more detail, but the subject is off topic here and Mr Born, having acted in an intellectually dishonest fashion before, is not worthy of further discussion.

Mr Born continues to be off topic, and characteristically sour. He made the false allegation, twice, that I had “refused” to provide data that he had not sent any request for, and that was already in our paper anyway. He has acted in bad faith. I do not propose to discuss this matter further with one who has proven himself to be intellectually dishonest. He is of course entitled to his opinion about our paper, for the opinion of a dishonest man is little regarded.