On Sep 11-12, 2006, KTH (Royal Institute of Technology) in Stockholm, Sweden hosted an international seminar on climate variability (seminar website here). The seminar had 16 speakers from 14 countries and was attended by 120 people. It was organized by Peter Stilbs and Fred Goldberg, who extended great hospitality to the presenters. Anders FlodstràÆà, President of KTH, agreed to the seminar and was an impressive figure as convener of the closing panel.
The seminar arose as one of a series of Pro-and-Con seminars sponsored by KTH. In this case, the balance of presenters and audience was non-IPCC. This was not through the fault of the organizers who made diligent efforts to obtain IPCC-types. However, in the end, von Storch, Bengtsson and Kallen ended up being the only "IPCC" presenters. Bert Bolin, former IPCC chairman, attended for part of the Monday session. (He refused to pay a conference registration of about $25 despite being asked for payment – I guess he’s used to expense accounts.)
The purpose of the seminar was not to present new results, but to summarize their views for a non-specialist audience. The following notes are not intended to be anything more than a rough impression and no slight is intended to those whose presentations are treated summarily.
On Monday norning, presentations were broadly speaking on reconstructions of climate history, with presentations by Fred Goldberg, Wjiborg Karlen, Bob Carter, Hans von Storch and myself.
Fred Goldberg is a material scientist, who has been an active "skeptic" in Sweden. He presented an account of historical information on the MWP and Little Ice Age. He showed some results on cloudiness thatI had not seen before, illustrated by some interesting paintings. (Fred travels to Svalbard every year and is familiar with the Arctic.)
Wjiborn Karlen is a prominent paleoclimatologist who has published dozens of peer-reviewed articles (curiously he’s joint on Moberg et al 2005). He presented information on variability in the Holocene. He showed the Briffa 2000 reconstruction – which, as I’ve pointed out here, is much influenced by the Yamal substitution. We chatted afterwards; he’s very concerned over the integrity of CRU temperature data and stated that no article involving Philip Jones could be relied on; I asked him if I could quote him on that and he said yes.
Bob Carter is an Australian geologist, who has professionally collected important deep sea cores from sediments offshore New Zealand showing climate variability in Deep Time. He presented on variability over Deep Time emphasizing that variability existed on every scale imaginable, showing what “trends” looked like over 10,000 years, 1000 years, 100 years and 10 years. He described the collection and interpretation of cores investigating northward flows of Antarctic water offshore New Zealand. He closed his presentation with an alarming quote from a reviewer for a grant who stated that grants should not be given to scientists who make public comments of the type that Carter has made. When I see the variability in Carter’s cores, in which centennial variability of the scale of the past century is routine, assertions that the variability over the past century requires anthropogenic influence seem rather over-confident. His PPT is online here.
Then moi. I explained how I got interested in the climate debate and how our present analysis evolved, beginning with the 2003 exchanges. Some aspects of our dialogue with Mann make more sense in this context. I presented a couple of new graphics — one showing the impact of one contaminated proxy on a von Storch-Zorita pseudoproxy network; one on Wahl and Ammann, but I’m finding that this sort of detail doesn’t play very well – not just with this sort of audience, but even highly specialized audiences.
Having said that, my pseudoproxy graphic did get understood immediately by von Storch. Von Storch and Zorita had sent me benchmark pseudoproxy results in the spring. They had reported last year (GRL) that Mannian PCA made "no difference" in a VZ pseudoproxy network in which the pseudoproxies were gridcell temperature plus white noise. In our Reply, we had argued that the pseudoproxy network was an irrelevant test for the impact of Mannian PCA on MBH proxies, but von Storch wasn’t convinced by the reply. A few months ago, they sent me a pseudoproxy run for me to reconcile. I replicated the VZ result on a pseudoproxy network of 55 series (their "region 1") agreeing that Mannian PCA did not have an impact on such a network. However, I showed a graphic illustrating the impact of the introduction of one synthetic nonclimatic hockey stick series on the network – Mannian PCA latched onto the nonclimatic HS and, under some circumstances, even flipped over the actual signal.
We had a nice chat in the afternoon – it was a beautiful sunny day in Stockholm. This is the third occasion tis year that we’ve co-presented: at the National Academy of Sciences, at the House Energy and Commerce Committee and now the KTH Seminar. He thought that my presentation was more relaxed than the previous presentations. He got the point of the new graphic instantly; he suggested that I publish it without mentioning bristlecones – as a purely mathematical exercise, although bristlecones will be the unmentioned elephant in the room. I guess one can be prudent in climate science from time to time.
Although my presentation was obviously very critical of the hockey stick, neither Bert Bolin nor Kallen had any questions or comments. The only critical question came from Bengtsson, who claimed that Mann’s error bars covered any problems. I replied that Mann’s error bars were meaningless as they had been calculated on calibration period residuals on an overfitted model and were “not worth the powder to blow them to hell”. Bengtsson did not pursue the matter.
Von Storch made a classroom-type presentation on detection and attribution. He had little invested in the presentation and pretty much mailed it in, but it was nice of him to show up. He challenged skeptics who thought that solar influence or any other influence was capable of explaining modern warming to do so in the context of a structured climate model — which seems fair enough to me although I wonder what sort of funding and support would be available for such an enterprise.
On Monday afternoon, presentations were broadly on the carbon cycle — Segalstat from Norway, Peter Stilbs of KTH, Jaworski from Poland, Sakalos from KTH and Richard Courtney from England.
I haven’t thought about the carbon cycle very much, but, thinking back, it was one of the first things that I wondered about and the IPCC answers are not necessarily obvious. The KTH scientists tended to come from a physical chemistry background and to approach carbon cycle issues from that viewpoint.
The connection of increased CO2 in the atmosphere to the burning of fossil fuels is "obvious" to climate scientists. 6 GT of CO2 are produced each year by burning fossil fuels and CO2 contents of the atmosphere increase by about 3 GT, with the natural system being unable to absorb the difference. This has gone on for the entire history of the Mauna Loa measurements commencing in the 1950s. Increased CO2 levels provide a plausible explanation for modern warming.
What troubles the physical chemists about this simple scenario is that "natural" CO2 flux in a given year is about 193 GT, so that the anthropogenic contribution is a relatively small contribution to the annual flux. Now it’s obviously conceivable that the natural sinks can’s accommodate the extra 5%, but there’s surely no harm in wondering about this topic. Further complicating the situation for physical chemists is that CO2 solubility in the oceans decreases with warmer oceans. Thus increased warmth would lead to increased atmospheric CO2 levels in a process that is readily understood in physical chemistry. Robert Essenhigh of Ohio State has written about this problem.
Here there’s an interesting connection with ice core results. In interglacial warm periods, temperature increases have led CO2 increases, rather than the opposite. The IPCC position is that the CO2 increases on Milankowitch scale are a feedback which are necessary to amplify the very small Milankowitch forcing. On the theory that the physical chemists are pushing towards, one would expect that CO2 levels in ice cores in (say) the warm Eemian interglacial would be similar to modern levels, rather than substantially less as observed in Vostok core. With this editorializing, on to the afternoon presentations.
Segalstat of Oslo University led off the afternoon with a general, and, as far as I can tell, uncontroversial exposition of the carbon cycle, but closed with the well-known cartoon showing the correlation between change in bathing suits from bloomers to thongs and global warming. At this point, Bert Bolin exploded at Segalstat (who apparently is a former student of Bolin’s) saying that he needed to read a text book. Bolin announced hat he was leaving the conference because it was such garbage. After some efforts to restore order, Bolin sat down for a few minutes and then left, still without paying his entrance fee.
Jaworski discussed problems measuring CO2 in ice cores. He’s been severely criticized and I do not pretend to be familiar with the controversy. He discussed problems in contamination of ice cores, pointg out, for example, downhole penetration of Pb and Zn. He said that ice cores showed sheeting fractures and that CO2 could easily be lost in such fractures. If so, measured CO2 levels in ice cores could have a downward bias (though presumably the higher-frequency results would have some meaning.) Afterwards, in conversation with Fred Goldberg, he thought that a useful experiment for a technical university would be to test CO2 behavior in ice core under various high pressure situations. That seemed like a sensible experiment to me — and one worth doing — although I’d be surprised if the results were dramatic. But you never know and the experiments are either worth doing or worth replicating.
This physical chemistry approach of KTH was well exemplified in an original paper by Sakalos, a young scientist at KTH. He showed that quartz sand had very large surface area and acted as a catalyst in the oxidation of methane. Because of the enormous amount of quartz sand, he argued that this previously undesrcribed effect had a material impact on the methane balance.
Richard Courtney, a well-known English skeptic, closed the day with an impassioned presentation about modeling the carbon cycle.
On Tuesday, there was discussion broadly speaking on climate models and forcings (two presentations by Willie Soon, one on behalf of Sallie Baliunas), Kallén of Sweden, Bengtsson, Marcel Leroux of France and Fred Singer, followed by a panel discussion (von Storch, Bengtsson, Carter and Singer) led by questioning from the President of KTH, who was very poised. Hans Erren discussed Arrhenius from a historical point of view.
Willie Soon emphasized the newness of quantitative measurements of the sun and the infancy of our knowledge. There is a common graphic synchronizing the various measurements of solar irradiance; Willis showed an interesting graphic with the unstitched measurements – which are all over the place. He discussed the differing estimnates of the amount of change in solar irradiance fron the Maunder Minimum to the present — ranging from 0.2% to 0.6%. The $64 question for people seeking to attribute climate change to solar variations is how relatively small changes in irradiance can impact changes in climate. One certainly emerged with the view that understanding of solar flux was, so to speak, in flux.
He made a second presentation on the impact of solar changes on climate. He’s interested in the impact of solar variabilty on the ITCZ – an extremely interesting issue IMHO. We chatted about this – he’d noticed my post on Kim Cobb’s corals, in which I interpreted her MWP results as showing northward ITCZ movement rather than a "cold" Pacific.
Kallen made a presentation on the Arctic Impact Assessment Study. He showed an interesting graphic in which warmth in the Arctic in the 1930s was recognized, but differentiated the post-1970 warming as being much broader. Comparing Arctic warmth in the 1930s with modern warmth was a recurrent theme. He didn’t discuss the Antarctic situation.
Bengtsson presented on climate models starting with their role in numerical weather prediction. Meteorologists take considerable pride in extending prediction accuracy from a couple of days to a week or so. He showed a graph showing the decay in correlation as weather prediction models moved out. I was going to return his confidence interval question by asking him what confidence interval he would place on a model once the correlation got to zero. However, I thought that too many presenters were showing up too often and held my peace.
Marcel Leroux presented on Mobile Polar Highs. While these seemed to be a plausible meteorological phenomenon, I didn’t understand what was controversial.about them or how they tied in to global warming controversies. Leroux has a new book and I’ll read it some time. Marcel Crok has talked to Leroux and his understanding is that Leroux’ view was that some parts of the Arctic were warming and others weren’t and that this could be explained through a change in wind circulation involving the Mobile Polar Highs. I’ll send Lerouz the dO18 graphic from Mount Logan which the proponents argue to signify a change in wind circulation.
Hans Erren made a very nice historical presentation on Arrhenius, providing strong evidence that Arrhenius had both fudged his data and misinterpreted effects related to salt prisms as evidence of CO2 absorption. Hans’ talk could definitely be developed into a publication in a history of science journal or even for a more general magazine like NWT or Scientific American.
Fred Singer discussed the recent CCSP report (we discussed the statistical appendix here recently), pointing out what appeared to be a pretty alarming statistical manipulation by Wigley and associates. I’ll take a look at this at some point., but, if true, is the type of manipulation that is unfortunately so common in this field.
A Swedish scientist made a presentation on impacts, but I missed most of this presentation.
The closing panel consisted of von Storch, Bengtsson, Carter and Singer, with Flodstrom, President of KTH, asking some questions and then turning questions over to the audience. Flodstrom handled this well. I won’t summarize this, other than to mention von Storch’s complaint about non-climate scientists butting into climate science – something that doesn’t happen in, say, biochemistry or particle physics. He pointed out that KTH was a technical university without any tradition in climate science, but still acted as host to a climate science seminar. Von Storch thought that this phenomenon deserved the interest of sociologists. It’s a fair question, which I think can be answered, but won’t comment on at this time.
Anyway, I found the conference interesting and appreciated the hospitality.