In the news this week, lots of agitation over some questionable science from an NGO wrongly cited by the IPCC, and a newspaper that caved to pressure.
The two journalists who originally broke the story “Amazongate”, Booker and North, were covered on WUWT last January. See links here and here. Now with new developments and a retraction by The Sunday Times, the controversy erupts anew.
Richard North writes on his EU Referendum blog:
Booker has taken on board the “Amazongate” developments in this week’s column. Interestingly, rather than me, it was Booker who suggested “going big” on the issue this week, his motivation in part being the intervention by George Monbiot, who has been his usual charmless self, parading the ugly face of warmism in all its triumphant ghastliness.
It is indeed getting ugly these days. It will likely get uglier as November elections in the USA approach. There’s a sense of panic afoot as some people know their window of opportunity is closing. Copenhagen failed, Cap and Trade in the US looks to be failed, Australia’s ETS is put on hold, and many other political objectives that are the result of an oversold set of actions are also unraveling.
Yes, the panic driven ugliness will get worse before it gets better.
The warmist community has gone into serious overdrive this week following the apology and correction in last week’s Sunday Times over its reporting last January of the IPCC scandal known as Amazongate.
The reason for all this? WWF, (World Wildlife Fund) which all you need to know. WWF is not peer reviewed science, it’s a billion dollar business with an agenda. When that business and it’s opinionated agenda driven output gets used in place of peer reviewed science, then all hope is lost for the integrity of science everywhere.
Let me remind everyone of the WWF sponsored report that led to the major 2035 glacier melt blunder by the IPCC. Texas State Climatologist John Nielsen-Gammon was the original finder of the error.
I covered the fallout here.
Last Friday, the WWF website posted a humiliating statement recognising the claim as ‘unsound’, and saying it ‘regrets any confusion caused’.
Dr Lal said: ‘We knew the WWF report with the 2035 date was “grey literature” [material not published in a peer-reviewed journal]. But it was never picked up by any of the authors in our working group, nor by any of the more than 500 external reviewers, by the governments to which it was sent, or by the final IPCC review editors.’
In fact, the 2035 melting date seems to have been plucked from thin air.
The WWF, in my view, is a poison pill for science. They should be avoided for any references in peer reviewed papers and in journalism.
In addition to the EuReferendum and Christopher Bookers column, we also have a fresh analysis by Willis Eschenbach on WUWT also well worth reading. This graph he produced sums up the entire issue succinctly: there’s no trend.
Booker at the Telegraph needs support now, more than ever before, please visit and comment on his article.
UPDATE: Shub Niggurath suggests that no peer reviewed science references existed in first and second order IPCC drafts:
More importantly, contrary to what many have suggested, it does not seem, that a statement was formulated assessing all available literature at the time. The sentence in question remained virtually unchanged through the drafts (except for the ‘drastic’ addition), it referred to the same WWF report through three different versions.
Well worth a visit to his site – Anthony
Spence – A global climate model is clearly in the categories ‘climate models’. It is not in the category ‘local things’.
Koutsoyiannis et al (2008) compared the real-world weather at a total of 8 stations with the weather in colocated clusters of one to four grid cells. This resolution, even over a period of 30 years is simply insifficient to extract the signal for the noise and I am not aware that anyone attampts to use single grid boxes in this way. I refer readers to Koutsoyiannis’s fellow Hydrologist, Z.W. Kundezewicz, writing in the Journal of Hydrology
“Koutsoyiannis et al. (2008, 2009) showed that the current generation of climate models
reproduced aspects of past climate (at the local scale) poorly, and concluded that “predictions”
based on these models are therefore unreliable. There are two flaws with this argument. First,
climate models are not designed to reproduce accurately local variations in climate from year to
year; they are designed to simulate broad features of the climate system and its variability. They
will not reproduce exactly observed past variability if they are not guided with accurate, timevarying
boundary conditions (such as time series of observed sea-surface temperatures), or if they
do not represent exactly all the processes influencing year-to-year variability. Climate models
which are driven with realistic sets of variable boundary conditions, such as observed sea-surface
temperatures, are much more able to reproduce observed patterns of climatic variability (Hurrell et
al., 2006). The second flaw, however, is more fundamental …..”
Next they expanded the study by adding more grid stations, presenting the results at the EGU (Poster presentations are not usually in the category ‘peer-reviewed’). Some may find this upscaling compelling evidence that models are useless for regional or even global projections (upscaling and downscaling are in the category ‘scaling’, however they …. oh never mind), and the extensive literature on the subject is now invalid. I reserve judgement. If K et are correct, then the Climate Model chapters of AR5 will be a lot shorter ;-).
Assuming Malhi et al [note the spelling] means this study, it does indeed state that
“Taking the ensemble of 23 IPCC models as a crude metric of probabilities, some intensification of dry seasons is about 80% probable in the southeast Amazon and Guyanas […etc ]. However I can find nowhere where this study or Nepstad rely on the results from a few runs or individual grid cells or small clusters, which would make Koutsoyiannis et al relevant. Perhaps Kim or Spence could point it out?
I do Kim the courtesy of assuming she has read the article she pronounces on.
Kim I think it is reasonable to assume that Dr Nepstad does not have an answer to these questions.
Your opinion is noted.
PC.
Brief Apology – I attributed the last post by Spence to Kim in error. My mistake.
Phil,
GCMs produce output of local climate through gridcells. It is clearly not a category error to ask whether they are meaningful. Changing the word “local climate” to “local things” does not a category change make.
I understand there is a disagreement over many of these aspects between scientists. That’s fine. But nevertheless the question: are these outputs meaningful? is still a valid, scientific question to ask. And these questions and their answers are published. As such they should not be ignored by the IPCC.
You say that climate models must be bounded by accurate sea surface temperatures to get local variations. (Dr Kundezewicz makes the same mistake here: sea surface temperatures are initial conditions, not boundary conditions, in the standard nomenclature of numerical modelling) But the results used by *Malhi 2007, cited by Dr Nepstad above to justify his views, have the same type of initialisation as the results used by Dr Koutsoyiannis. Dr Malhi does not use the global data but selects gridcells over the Amazon and comments on them. As noted, an ensemble cannot “fix” the problems with precipitation as the results are not “noisy”, but biased.
Dr Kundezewicz also notes that climate models are not well suited to producing local climate: yet another scientist who outlines (just as you have above) that the methodology by Malhi 2007 is fundamentally flawed. Thank you for continuing to find people who show that you cannot derive local (or regional) climate from the IPCC AR4 runs in the way Malhi did. You underline my point very clearly.
As for the analysis of the US 48 contiguous states, it has not been published at this time. Neither, of course, have Gavin’s criticisms of the work. This does not make either of them wrong. As noted, they decided not to submit some of the results (as they showed nothing new, and were not interesting enough to publish). However, it is a part of an ongoing project and I understand there are intentions to submit further articles to journals in the future on this topic.
As for IPCC AR5, it is structured in the same way as AR4, so I fully expect it to present a narrow viewpoint, rather than a representation of a wide range of scientific views. I may be wrong: time will tell.
I do not understand your commentary on the Malhi study though. Clearly, Malhi has looked at groups of individual grid cells selected over the Amazon region. There are diagrams of them in the article (which I have read, thank you). How else do you think the conclusions were reached? These are local climate variations. The ensemble does not help precipitation issues, as Koutsoyiannis et al clearly showed the precipitation results are biased, and averaging cannot remove bias (although it can confer a false sense of security by reducing variability).
In summary: Dr Nepstad relies on Malhi 2007 to justify his claims that the IPCC is correct. Malhi 2007 selects gridcells over the Amazon rainforest from IPCC AR4 model runs to determine future rainfall characteristics – the same runs that Dr Koutsoyiannis has demonstrated do not reflect local climate. In addition, many other scientists – including Dr Schmidt and Dr Kundezewicz – agree that these model runs cannot accurately capture local climate, as you have noted.
* thanks for catching the spelling error. I noted Dr Nepstad had an incorrect spelling above, and knew there should be a letter “h” in it – I put it in the wrong place. Now, hopefully, corrected.
Last try then I give up:
Koutsoyiannis (is he a Professor or a Doctor of Hydrology?) ‘demonstrated’ something that was already well-known: viz.
Can GCMs predict the temperature and precipitation for my home?
No. There are often large variation in the temperature and precipitation statistics over short distances because the local climatic characteristics are affected by the local geography. The GCMs are designed to describe the most important large-scale features of the climate, such as the energy flow, the circulation, and the temperature in a grid-box volume (through physical laws of thermodynamics, the dynamics, and the ideal gas laws). A typical grid-box may have a horizontal area of ~100×100 km2, but the size has tended to reduce over the years as computers have increased in speed. The shape of the landscape (the details of mountains, coastline etc.) used in the models reflect the spatial resolution, hence the model will not have sufficient detail to describe local climate variation associated with local geographical features (e.g. mountains, valleys, lakes, etc.).
So GCMs do not correlate well with observations at the local level. Woop-de-doo. Let’s define ‘local’ as 1-4 grid cells or (generously) < 250x250km. Such discrepencies are irrelevant to and unmentioned by Malhi or Nepstad, despite claims above. (Of course they looked at 'groups of individual gridcells', the whole planet is 'some gridcells'!). They used regional projections. Any sensible definition of 'regional' in this context must mean millions of km2. The Amazon rainforest covers about 8,000,000 km2. I am simply not convinced that K's local findings can be extrapolated up to the regional scale referred to by Nepstad and Malhi.
TTFN.
Phil,
I can see why you’re giving up. I’ve told you many, many times that the Koutsoyiannis work was extended to a region of 8 million sq km. And each time you bring back the same criticism: that the analysis only applies to single gridcells. Unfortunately, closing your eyes will not make this inconvenient fact go away for you.
Your initial comments and criticisms were quite constructive, but they are now becoming a mixture of repeated criticisms that have been addressed (such as the single gridcell criticism) and new questions which are simply odd:
He is a professor and a doctor. At least, last time I checked, having tenure did not result in the retraction of a previously earned doctorate.
Yes, we have discussed this. This is *one* of the results of Koutsoyiannis’ work. One that you have provided considerable evidence to support.
But this is not the only result. And your decision to wilfully ignore the second result – that spatial averaging to regional analysis makes the results worse – means that you miss the point.
Imagine this conversation:
Spence. “I have a study showing clouds are white and the sky is blue”
Phil. “No, the sky is pink.”
Spence. “My study shows that the sky is blue. You are wrong.”
Phil. “No, your study shows the clouds are white. This is irrelevant to my observation that the sky is pink”
Spence. “…”
Back to your post:
Firstly: Koutsoyiannis’ study covered both single points and a complete region of 8 million sq km.
Malhi 2007 states.. “the forest of Amazonia covers 5.4 million sq km”. Somehow you have inflated that to 8 million sq km. Because we are referring to the Malhi study, I will use Dr Malhi’s figures.
As you can see, 5.4 million square km lies between the data points of Koutsoyiannis study, from single grid cells to 8 million sq km. Therefore, we are extrapolating nothing. Indeed, in numerical analysis we refer to this as “interpolation”.
And you are deeply unfair to Dr. Malhi. The analysis by Koutsoyiannis was completed some time after the Malhi paper was published. Of course it is unmentioned!!! Unless they have a time machine, I can’t see how it could be mentioned. Because the Koutsoyiannis work post dates the Malhi work, we must assess the implications for ourselves.
Unsurprisingly, we may make different interpretations and come to different conclusions. However, to simply pretend the regional analysis conducted by the Itia group does not exist, after being told so many times, is an astonishingly weak viewpoint.
I can’t stop you from holding it though. That’s your choice.
To James Sexton.
I like your style. I like it a lot.
I must comment one thing that Nepstad writes, even if it doesn’t seem like he will be return to this thread:
“If those future drought episodes are more severe than in the past, then lots of damage from drought and fire could ensure.”
Is anyone really questioning this?
It would be like questioning a statement like: “If the world explodes, we all die.”, or: “If the Amazon burns to the ground, then there is no more Amazon.”
This is clearly not what Booker, North etc. have been questioning.
The above comment, and much of the other comments from AGW-proponents on this subject, are mere smoke screens.
Since good science is made up of carefully gathered details, reports about science (like the IPCC reports, but also those in the mainstream media) should be true to those details and criticism concerning these details should NOT be met by sweeping, non quantative, loosely related replies that only serve to obscure these details further.
AGW-proponents seem to be fond of using the “Would you risk standing in the middle of a highway?”-analogy, what they don’t seem to grasp is that first you must know that where you’re standing really is a highway, that it actually is up and running and that there actually is a risk of being hit by a vehicle.
What they’re, including Nepstad, actually saying is:
“Because there exist highways that are populated by potentially lethal vehicles in motion, you should not stand, sit or walk anywhere on this earth, because the place where you choose to tread could be a highway!”
“quantative” = “quantitative”