From	Richard Alley	Date	Sat, 11 Mar 2006 15:27:19 -0500 (EST)
To	Eystein Jansen, Fortunat Joos, Jonathan T. Overpeck, Keith Briffa, Bette Otto-Bleisner, Phil Jones, Ricardo Villalba
CC	Martin Manning, Susan Solomon
Subject	NRC and IPCC millennial temperatures
Friends in the IPCC WG1 AR4-- My impression is that, for good reasons, the US NRC panel looking at the record of temperatures over the last millennium or two is not going to strongly endorse the ability of proxies to detect warming above the level of a millennium ago, and that a careful re-examination of the Chapter 6 wording and its representation in the TS and SPM would be wise. Some of you have seen some of the discussion that follows, in some of the rapid-fire emails over the last day or two, but I'd like to clarify a little. Please note that I am NOT on the NRC committee, do not speak for them, and have no "inside" knowledge of what they are doing. I was asked to testify to them, and I heard remarks from some other speakers and questions from the committee in public forum. I did NOT represent the IPCC to the committee, either; I stated that although I was proud to be participating with the IPCC, I absolutely was not speaking for, representing, or presaging anything in the IPCC. (I was, however, favorably quite impressed with the NRC committee and their efforts.) Someone else may have a different impression of what went on; this is mine. Among the presentations, involving borehole temperatures, corals, glaciers and ice cores, and historical records, that which to me seemed to interest the committee most was from Rosanne d'Arrigo, who reported (among many other things) on a just-published study in which northern tree-ring sites were revisited and updated, and in which many of those sites failed to track the recent warming documented instrumentally. She did not make a big deal out of this, but several of the questions afterward from the committee focused on this "divergence" problem. (And to note, Rosanne did not discover the divergence problem, which has been around and discussed for a while; her testimony, including the recent large effort to update some tree-ring records, stirred interest from some committee members.) I would also note that one of the committee members was asking each presenter whether the presenter believed that temperatures could be reconstructed for 1000 years ago within 0.5 C, and that the presenters were answering with some qualified version of "no". My guess is that the NRC committee will put these things together, find some papers on ozone damage and CO2 fertilization, consider Rosanne's statement that the preferred temperature-sensitive trees are rare and in restricted places (and thus that a prolonged warming could easily move those trees out of the sensitive band), and conclude that tree-ring reconstructions include larger errors than are returned by any of the formal statistics from calibration or aggregation of records, and thus that there is less confidence than previously believed in the relative warmth of recent versus Medieval times. I also consider it possible that they will point out the difficulty of using a composite temperature history consisting of proxy and instrumental data if some of the proxy data do not track the more-recent part of the instrumental data. The IPCC must be the IPCC, not the NRC. But, if the IPCC and NRC look very different, there will be much comment, and we will have to be very sure. More importantly, I believe that real issues are raised here, and that better discussion of this should be included in chapter 6, and probably brought forward at least into the TS. I know I'm not in chapter 6, I know I'm not a tree-ring expert, and I know I'm sticking my nose in where it might not belong or be welcome. But the flurry of emails in the last couple of days has not convinced me that this one can be ignored; indeed, I am more convinced that there exist issues that the IPCC must discuss more thoroughly. My impression of the status (and my thoughts about what chapter 6 might say) from a whole lot of quick reading, your emails, and the testimony and questions I heard, is along the lines of: --> The TAR highlighted a temperature history composited from multi-proxy paleoclimatic indicators plus the instrumental record, showing anomalous recent warmth, with the recent warmth emerging well above the 95% confidence interval for the last millennium. --> The multi-proxy paleoclimatic indicators reflect tree-ring results more than any other source. --> Tree-ring records are responsive to many factors, and great care and effort go into isolating the temperature signal from other signals. --> Tree-ring data, in common with essentially all paleoclimatic data, are not collected in a continually updated "operational" fashion analogous to that used for meteorological data, so the data sets end at different times; data used in the multi-proxy reconstructions cited in the TAR ended between the 1990s and the 1940s. This difficulty motivated the need to include instrumental as well as proxy data in the reconstructions. --> In those data, there was some suggestion of non-temperature influences on the tree-ring reconstructions; in particular, some of the most-recent records did not record the full amplitude of the instrumental warming. This has come to be known as the "divergence" issue. --> Much research has been conducted since the TAR, and additional evidence of divergence has emerged in some records, causing some aggregated reconstructions from proxy records to show less warming than does the instrumental record. --> There are many hypotheses for non-temperature influences on tree-ring records, including: (i) recent damage (as by ozone); (ii) recent fertilization (as by CO2); and (iii) decreasing sensitivity of tree-ring growth to temperature with increasing temperature (once it's warm enough, the trees are primarily responsive to other things). The nature of these and their timing relative to the interval in which tree-ring data were calibrated to instrumental records would control the effects on climate reconstructions. In general: (i) would mean that recent warmth is underestimated but warmth from a millennium ago is not; (ii) would mean that recent warmth is overestimated but warmth from a millennium ago is not; and (iii) would mean that both recent warmth and warmth from a millennium ago are underestimated. --> Various arguments have been advanced to support (i), (ii), or (iii), with many workers in the field favoring (i). Nonetheless, further characterizing recent non-temperature influences on tree-ring growth remains an open research question, and no broad consensus has emerged on (i), (ii), (iii), or something else. --> These considerations do not affect the conclusion that recent warmth is anomalous over the last few centuries; the strong correlations of the proxy data with temperature over the instrumental record, and the strong tree-ring signals, are evident. --> These considerations do not affect the best estimate that recent warmth is greater than that of a millennium ago; the central estimate from proxy data of latter-twentieth-century warmth is still above that of a millennium ago, with greater spatial conherence recently in the signal. --> These considerations do somewhat affect the confidence that can be attached to the best estimate of recent warmth versus that of a millennium ago. If the paleoclimatic data could be confidently be interpreted as paleotemperatures, then joining the paleoclimatic and instrumental records would be appropriate, and the recent warmth would clearly be anomalous over the last millennium and beyond. By demonstrating that some tree-ring series chosen for temperature sensitivity are not fully reflecting temperature changes, the divergence issue widens the error bars and so reduces confidence in the comparison between recent and earlier warmth. --Richard Richard B. Alley Evan Pugh Professor Department of Geosciences, and Earth and Environmental Systems Institute The Pennsylvania State University 517 Deike Building University Park, PA 16802, USA ph. 814-863-1700 fax 814-863-7823 email rba6@psu.edu