HYPERLINK "http://www.climateaudit.org/?p=962" \o "Permanent Link to Day Two at AGU" Day Two at AGU 

http://www.climateaudit.org/?p=962#more-962

By Steve McIntyre

It is impossible to convey the overwhelming number of papers and presentations here. Through the week, my notes have invariably deteriorated. By the time you get home, they are unintelligible. I’ m going to diarize them a little – so I don’t forget totally and to share a bit of the experience. (And since CA readers have in effect paid for the trip, I’ll try to report as well as make my own presentation.)

One comment which should be re-assuring to many readers. There are many younger scientists starting to push their way into positions where they are evaluating the climate of the last 1000 years and I feel quite confident that a reasonable view on the matter will emerge within whatever material is available. Andy Bunn and Andrea Lloyd have a NSF contract to to do a big update of tree ring sites. Rob Wilson is trying to resolve the divergence problem on new sites, without using the same sites. Julie Richey is doing high-resolution analysis of new cores and has no patience for endlessly re-using the same data. Alicia Newton’s high resolution core in the Western Equatorial Pacific is an excellent contribution. Who knows what the results will actually be?

BTW the exchange between Mann and myself was reviewed by a third-party blogger  HYPERLINK "http://geology.about.com/b/" here. The blogger was impressed by Mann’s newest multivariate method - I suspect (in fact I’m sure) that he had no idea what the new method was; if Jean S were impressed, that would mean something. But all the new method is going to be is a new method of making a linear weighting of the proxies. If there’s a “signal”, the precise method of weighting really shouldn’t matter very much. 

I get quite a different personal reaction from the young scientists. Both Bunn and Wilson were very friendly and interested in what I was planning to do. In reverse, I was glad to see that both of them are progressing in their field. Bunn has moved on from a post-doc position to a tenure-track position. Rob seems to have established himself at the University of Edinburgh. In a way, despite my age, my perspective on the field is more like that of a young grad student or post-doc – everything is still fresh (except maybe principal components which I’m bored with). When you do one thing too long, you get stale. Anyway on to my diary.

I started off this morning going to some posters on ocean sediments, which I’ve been following. I can do little more than list some features. PP21A-1663 was a high-resolution core MD99-2275 off north Iceland at 2-5 year resolution, using a new proxy IP_25 as a sea ice proxy with peaks in 1776, 16888, 1638 etc.

PP21A-1668, Site 1094 in the Antarctic, showed a Holocene maximum. #1672 showed some trees from the Pliocene Arctic – similar to something I posted up before. I browsed many other posters.

I went to the two hurricane presentations. Kerry Emanuel was supposed to present, but was a no-show and Mark De Maria filled in, discussing improvements in forecast skill of hurricane track and intensity – track forecasts have improved much more than intensity forecasts.

Then Peter Webster of Webster and Curry presented on the tole of hurricanes in the global heat system. He observed tat the number of hurricanes and tropical storms per year (81 and 46) was very stable and proposed that this meant they fit into some sort of heat balance. The question was then: why 80 and not 8 or 800? This seemed very plausible – in fact, I was surprised that it was being presented as a novelty. For some reason, I’d always assumed that storms would have some kind of Maxwell-Boltzmann distribution. He then compared energy totals for hurricanes to thunderstorm estimates of Riehl-Malkus in a back-of-envelope calculation (a highly sensible back-of-envelope calculation – I’m not at all critical of the calculation).

I then went over to session PP22A on South American proxies in the Holocene. I missed a paper by Rodbell that I was interested because it was on at the same time as Webster. PP22A-04 reviewed various LGM proxies in the the SH. PP22A-05 by a climatologist, said that reconstructions talking about ITCZ movements tended to confuse the ITCZ – an oceanic phenomenon, with the South American monsoon, a different process. Kerry Cook said that in the LGM the monsoon was delayed by 2-3 months because of cold Atlantic SSTs, resulting in a longer dry season and reduced rain forest. Kaplan (PP22A-06) discussed deglaciation in S Patagonia, especially Torre del Paine. Makou (PP22A-07) reported on ODP 201-1228D on cholesterols dinestrol as proxies. Turc1 PP22A-08 reported on increasing lake levels in N Brazil through the Holocene.

I started off the afternoon at GC23C – Terrestrial Records of Holocene Climate Change from the Tropics. Lonnie Thompson presented GC23-02 (he also gave a lecture the day before which I missed). It was pretty standard stuff – after his speech, he swept out of the room, unlike almost any other participant in a session. HE showed the usual photos of receding glaciers. He said that Quelccaya went back to 315 - this must be in the new cores as the earlier cores went back to 470. Two new cores were drilled in 2003.) He showed some chemistry from Quelccaya – he said that chemistry for the original cores was not available because they were then unable to return the cores frozen – only in bottles. That doesn’t sound like an explanation to me – why can’t you do chemical analysis on water? I wonder if the bottles were lost. HE showed some graphics from Coropuna only 70 km from the ocean. It was drilled in 2003 as well – 148 m; of which 33 m to 1975; 55 to 1940; 80 m to 1900 – to give you some idea of the exponential shortening of the layers. HE showed some graphics from Dasuopu talking about the 1796 monsoon – just like his 1997 article. HE showed the Quelccaya plants dated to 5200 BP. He showed a picture of an outcrop with a simple picture of a geologic cross-section (hurray)with clay on the bottom and gravels on the top, radio carbon dates increasing from 5152 to 6703. No discussion of reservoir effect – something that every other paleoclimatologist discusses almost first thing. He showed Rodbell’s diagram of Pallcacocha – the one which I showed last summer on CA in discussing Quelccaya. HE said that Rodbell “interprets this as the approach of the glacier” whereas Rodbell had said that it marked the initiation of the glacier. He showed some other series also showing change around 5200 BP – he didn’t mention that this date is more or less what Porter and Karlen had proposed as early as 1976 for the start of the Neoglacial/Atlantic period. His most recent drilling is at Naimonna-ny in Tibet. Someone asked him about whether any glaciers had disappeared in the MWP – we know that Polissar et al said that Venezuela glaciers had disappeared. HE went into an explanation of temperature and precipitation as both being implicated in the dO188 signal, but regardless claimed that the 20^th century was unique as a signal. HE said that there was no correlation between thickness and isotopes, which he took to exclude the amount effect as dominating in the isotopic record. He said that there was modern melting at Quelccaya but no comparable melting in the past; but then said that melting sometimes occurred at low temperatures due to intense radiation.

The next presenter, Fritz, gave a nice presentation on high Altiplano lakes, stating categorically that dO188 was a precipitation signal, providing very convincing arguments. She studied Lago Lagunillas and L Umayo, near L Titicaca. The sediments had 5-year resolution. The lake water dO18 was connected to precipitation and it, in turn, correlated to Quelccaya isotopes. She compared the record to the Bond et al Hematite stained grain record, but the comparison was unconvincing to me.

Then a couple of presentations on N-S movements of the ITCZ and their impact on African equatorial lakes. Russell reported on Mg/Ca in Lake Edward, saying that Mg/Ca ratios increased with salinity.

Then an interesting presentation by Jessica Conroy (coauthor Overpeck) about a lake record in the Galapagos and the drought in the American SW. Her record was El Junco Crater lake, a 3.5 m sediment core and 3 short cores for the interface; she measured diatoms every 0.5 cm , counting 300. She did a Pb210 age model for 120 years (20 cm) and found that F saxomica had a correlation of 0.83 to Galapagos SST. There were no diatoms prior to 770 AD – which she attributed to drying. Her record showed medieval “cooling” - now Galapagos is an upwelling site and % G Bulloides (cold water) offshore Oman is taken as an index of global warming, so I’m not sure why upwelling of cold water offshore Oman is a a sign of warming, but upwelling in Galapagos is different. Of the two medieval megadroughts, one corresponded to a “cool “ Galapagos period, but they other didn’t. She noted a warm record in Keigwin’s Sargasso Sea SST record and associated warm Atlantic SSTs with the medieval megadrought (warm Atlantic SSTs are presumably the MWP). Her diatoms had a pronounced modern warming trend, but another species didn’t. Fertilization was mentioned as something that needed to be kept in mind. She mentioned Kim Cobb’s “cool” medieval Pacific – I’ve discussed her results in the context as possibly being no more than a slight northward movement of the ITCZ,

I then went over to OS23E to catch a presentation coauthored by Hughes (but the key player was N Graham) on the California medieval droughts. Graham - “some really extreme things happened not very long ago”, “much more severe than anything in modern record”. I saw Rob Wilson during this presentation who was very cheerful. Graham listed a variety of proxy records – Crater Lake char, San Francisco Bay salinity, Santa Barbara sediments and an unspecified bristlecone record (going down in the 20th). Graham then proceeded to analyse Mono Lake, a closed lake east of the Sierra which received 85% of its inflow from Sierra snow. Stine had found relict trees there in two populations from the MWP; the trees were Jeffrey pine which dO1 not tolerate submergence and which would have grown beside a much retrenched lake in the MWP. He modeled Merced River flow as being 25-40% less than modern flow from tree ring drought index reconstruction- he didn’t specify which tree rings he used for the drought reconstruction in his presentation. He identified two deep medieval droughts. He then used the river inflow reconstruction to reconstruct Mono Lake levels and produced a graph of lake levels that pretty much replicated Stine’s results in a graphical way – he said, at the first go without any tuning. Looked plausible to me. He then asked what circulation changes could cause such results – drought in the Sierras has for many years been associated with jet stream further to north (causing moisture in Alaska). He linked the MWP drought to warmth in the Mindanao SST (Stott); cool Santa Barbara SST (upwelling), and Peru river discharge. He cloased by re-emphasizing the scale of the events - “much more severe than anything in the modern record” and that the MWP events were anomalous even in a 4000 year context. So when coauthor Hughes wonders with Diaz about Was There A Medieval Warm Period?, he might discuss the matter with his other coauthor Graham. I’m not sure that there’s anything in Graham’s presentation that I would disagree with.

I then went over to Andy Bunn’s presentation at B24A. I almost missed this as I wasn’t looking at these sessions, but I ran into Rob Wilson at the other session. The room was too small and Rob went back to the other session which was too bad as he would have liked Bunn’s presentation. Bunn is a really sharp young ecologist and tree ring analyst. I mentioned before that he and Andrea Lloyd have a grant to update chronologies and people can rest assured that this program is in good hands. The results will be what they will be, but there won’t be any cherrypicking or ex post deciding about what sites were temperature-sensitive (and archived) and what sites were not and discarded (a la you-know-who). Anyway, Bunn talked first about “greening” and “browning” in the boreal forest as indicated by NDVI-AVHRP imagery. He compared these indices to ring width chrnologies and found that there were good correlations between greening and high ring width and browning and low ring width indices; he analyzed CRU half-degree interpolated cells and station data. He showed ak036 – a negative responder – against station data; somehow he was able to find out that the station was not in CRU records. He showed the difference between a greening chronology – larch 65N, 169W and a browning chronology ak036. He analysed a lot of sites (hurray for studying real populations and not cherry-picked subsets). He showed some nice bar charts showing the proportion of positive and negative responders by species. Larch mostly had postive responders, but white spruce, the staple of many chronologies, was half and half. He showed differences between the pre-1950 and post-1950 patterns – with decent correlations pre-1950 and negative correlations post-1950: the Divergence Problem. He said that these results were causing him “great stress”, (as they do for Rob Wilson). The stress for him was not so much that the divergence might be caused by negative responders due to precipitation-temperature interactions, but by phenotype plasticity.

I closed by attending a session on the difference between evapotranspiration and canopy transpiration in boreal forests, It didn’t relate very much to anything that I was interested, but it was nice to listen to someone talk knowledgeably about his field without once using the word “unprecedented”.