cc: c.goodess@uea.ac.uk,k.briffa@uea.ac.uk,t.osborn@uea.ac.uk date: Thu, 25 Sep 2003 14:13:58 +0100 from: Phil Jones subject: Fwd: Re: new scientist to: m.hulme@uea.ac.uk,h.j.schellnhuber@uea.ac.uk Dear All, The issue has moved on a little. The editor of NS will not accept another piece, only a letter, which Stefan Rahmstorf has drafted. I've not had a chance to look at it, but if anyone wants to join Stefan can they get in touch with him directly. I am going to sit this one out. I am a little alarmed by Mike Mann at times, but his comments are only ever in this friendly email context. Cheers Phil X-Sender: mem6u@multiproxy.evsc.virginia.edu X-Mailer: QUALCOMM Windows Eudora Version 5.2.1 Date: Thu, 25 Sep 2003 08:25:28 -0400 To: Stefan Rahmstorf From: "Michael E. Mann" Subject: Re: new scientist Cc: Gavin Schmidt , cindy@stopesso.com, André Berger , Phil Jones , Maraun , mann@virginia.edu Stefan, It looks great to me, I wouldn't change anything except perhaps, the final clause of sentence #1: which received a disproportionate amount of media coverage due to the unfounded claims the authors made in their press releases ("Only about a third of the warming over the past century should be attributed to man"). to which received a disproportionate amount of media coverage due to the unfounded claims the authors made in their press releases that "Only about a third of the warming over the past century should be attributed to man". Your point about the problems in using a regression of empirical estimates of response against forcing is an important one. The main problem here is that the authors supposed "global temperature" estimate is nothing of the sort. I actually did some research into this issue and here are my comments: Veizer's estimates are almost certainly not representative of the quantity claimed by Veizer (i.e., tropical mean sea surface temperature). Going back to Veizer's original (1999) "Chemical Geology" paper describing the data, I found some troubling issues in the description of the data. The data were collected from a highly irregular and inhomogenous spatial network of locations over the modern continents. The authors argue, based on paleogeographic reconstructions, that "most of the data come from the tropics". That is a disturbingly poor basis on which to define a composite of the data as a supposed estimate of tropical mean SST! No account seems to have been taken for whether or not a simple mean over the available sites is likely to represent a representative areal average of the tropical oceans (it can easily be shown that a similar random sampling of site-based SST measurements from the modern instrumental data base will generally give a substantially biased estimate of the true tropical mean SST variations). Climate scientists take great pains to insure that they average a set of site measurements in such as way that a meaningful areal (e.g. tropical, Northern Hemisphere, or global) average can be computed. A tropical SST estimate based mostly on tropical Pacific instrumental data, for example, would overly emphasize SST variations related to ENSO, and give a biased picture of global tropical SST. There is no evidence in anything I've read in Veizer' papers, that care was taken to insure a meaningful spatial mean estimate of tropical SST. Equally problematic is the changing distribution of sites and data sources over time, which may considerably bias the record. Veizer himself (2000) notes, in fact, that the Neogene estimates may be overly dominated by data from the North Pacific. These are all possible reasons for why the Veizer estimates may not be reliable estimates of the quantity (tropical mean SST) claimed. This may contribute to why they do not show good agreement with other (e.g. glacial) evidence (i.e., Figure 2A vs Figure 2C) even after correcting for the Ph effects, and thus cannot be used to infer (as in Shaviv and Veizer) an estimate of the sensitivity of the global climate to co2. In fact any estimate of sensitivity from a regression analysis will in general underestimate the sensitivity (unless the forcing and response are self-consistently estimated, as in a forced model simulation). This has to do with the fact that the uncertainties in the forcing and response are independent, and while the uncertainties in the numerator of the expression used to derive the sensitivity from the data covariances cancel, the uncertainty in the forcing series artificially increases the estimated variance in the forcing series, which increases the dominator. I discussed this issue at some length in this paper: Waple, A., Mann, M.E., Bradley, R.S., [1]Long-term Patterns of Solar Irradiance Forcing in Model Experiments and Proxy-based Surface Temperature Reconstructions, Climate Dynamics, 18, 563-578, 2002. available here: [2]ftp://holocene.evsc.virginia.edu/pub/mann/WMB2002.pdf cheers, mike At 01:37 PM 9/25/2003 +0200, Stefan Rahmstorf wrote: Hi everyone, I'm thinking of sending the following letter to New Scientist. Please check critically what I say to make sure it stands up under fire. Your comments will be most welcome. Stefan --- Stott claims that the paper by Shaviv & Veizer is important science that did not get enough attention from media and policy makers. The opposite is true: it is a paper of little scientific credibility, which received a disproportionate amount of media coverage due to the unfounded claims the authors made in their press releases (Only about a third of the warming over the past century should be attributed to man). Shaviv and Veizer claim to have found a correlation between cosmic ray flux and temperature. Even if we accept their (questionable) data, it should be noted that this correlation was constructed by arbitrarily stretching the time scale to shift the maxima of cosmic ray flux by up to 20 million years, to make them coincide with temperature minima. The unadulterated data show no significant correlation (we checked this). Shaviv and Veizer then proceed to estimate the climate sensitivity to a doubling of CO_2 concentration through regression analysis, which for a number of reasons is not possible. If it were, far better data could be used for this analysis: the Antarctic ice core data, which are more accurate, show variations on more relevant time scales (not tens of millions of years) and closer to present CO_2 levels, and apply to the present-day configuration of continents. This would yield a climate sensitivity exceeding 10ºC, but no climatologist has suggested this is a viable method. Climatologists agree that doubling CO_2 concentration would heat global climate by ~2-4ºC, not because this is a hegemonic myth but simply because this conclusion is based on sound science: the known radiative properties of CO_2 and an understanding of the key physical feedbacks in the climate system. -- Stefan Rahmstorf Potsdam Institute for Climate Impact Research (PIK) For contact details, reprints, movies & general infos see: [3]http://www.pik-potsdam.de/~stefan ______________________________________________________________ Professor Michael E. Mann Department of Environmental Sciences, Clark Hall University of Virginia Charlottesville, VA 22903 _______________________________________________________________________ e-mail: mann@virginia.edu Phone: (434) 924-7770 FAX: (434) 982-2137 [4]http://www.evsc.virginia.edu/faculty/people/mann.shtml Prof. Phil Jones Climatic Research Unit Telephone +44 (0) 1603 592090 School of Environmental Sciences Fax +44 (0) 1603 507784 University of East Anglia Norwich Email p.jones@uea.ac.uk NR4 7TJ UK ----------------------------------------------------------------------------