date: Wed, 28 Feb 2001 12:28:29 -0500 from: "Michael E. Mann" subject: letter to Science to: tom@ocean.tamu.edu, td@gfdl.gov, hpollack@geo.lsa.umich.edu, mhughes@ltrr.arizona.edu, mann@virginia.edu, rbradley@geo.umass.edu, p.jones@uea.ac.uk, k.briffa@uea.ac.uk Dear Colleagues, Below is a draft of a short letter to Science that Tom Crowley and I have put together, after discussing w/ Phil, Ray, and Malcolm. We feel that a reply to Broecker's recent "Perspectives" piece is warranted to correct several misconceptions that Wally unfortunately chose to perpetuate (attached as an html file FYI). We have been given encouragement to submit this by Julia Uppenbrink at Science. We are working under a very tight timeline owing to Tom's travel schedule (leaves on an extended travel on friday) so we would greatly appreciate it if you could respond ASAP w/ comments, suggestions, etc. Please note that we are currently near the length limitations (and probably shouldn't include more than 15 references) so we're looking to sharpen and hone, but not lengthen the piece at this point. Thanks in advance for your feedback, mike _________________________________________ Medieval Warming Redux In a recent "Perspectives" opinion piece, W. Broecker suggests that the "hockey stick" reconstruction of climate change over the past 1000 years - with extreme warming only in the late 20th century - is incorrect, and that the so-called "Medieval Warm Period" was at least as warm as the 20th century and due to oscillations in the thermohaline circulation. To reach this conclusion, Dr. Broecker rejects traditional empirical "proxy" climate indicators of past climate (e.g. tree ring, ice core, coral, and long historical documentary records) that are the foundation of a number of hemispheric reconstructions, as well as our current best physical understanding of the factors controlling climate at century-to-millennial timescales. We disagree with Broecker on several major points: (1) It cannot reasonably be argued that the Middle Ages were as warm as the 20th century at global or hemispheric scales. Although regional warmth during the Middle Ages may have sometimes been significantly greater than present, four different hemispheric-scale reconstructions (Jones, Mann, Briffa, Crowley) have been completed for the last 1000 years -- all of them showing warmth in the Middle Ages that is either no warmer or significantly less than mid-20th century warmth. This is because it has been known for a quarter of a century that the timing of warmth during the Middle Ages was significantly different in different regions (Lamb, Dansgaard, Hughes). Failure to take this observation into account can lead to serious errors in the inference of hemispheric temperature trends. Although one analysis of heat flow measurements suggests warmer temperatures than the surface proxies during the Middle Ages (Huang and Pollack, GRL. 1997), the considerable sensitivity of the resulting trends to a priori statistical assumptions has lead borehole researchers to restrict their attention to the more reliably interpretable temperature fluctuations during the past five centuries (Huang and Pollack, Nature). Our conclusion is also supported by measurements from tropical glaciers indicating an unprecedented level of recent warming with respect to the last 1,000-2,000 years (Thompson). (2) High-resolution proxy climate records which form the foundation of recent hemispheric temperature reconstructions are far more reliable indicators of century-to-millennial scale climate variability than is implied by Broecker. The potential limitations in interpreting long-term climate change from proxy indicators such as tree rings, have been long recognized by dendroclimatologists (e.g., Cook "segment curse" paper) and are almost always taken into account in framing interpretations of long-term trends. For example, Mann et al (1999) verified that a significant subset of multiple-millennial length tree ring and ice core proxy climate indicators used to reconstruct the trend over the past millennium passed rigorous statistical tests for fidelity at the millennial timescale, and that the basic attributes of the hemispheric reconstruction using more recent non-tree ring proxies available over the past few centuries yielded essentially the same result as that based on both tree ring and non-tree ring based information (Mann et al, Earth Interactions, 2000). Several independent reconstructions (Jones et al and Crowley and Lowery ), using a wide variety of proxy climate indicators and different statistical approaches, yield similar hemispheric temperature trends. Even the centennial-scale changes within the so-called "Little Ice Age" of the 15th-19th centuries are largely in agreement. Furthermore these centennial changes have been shown to be in "agreement" , rather than "in opposition" (as argued by Broecker) with evidence from alpine glacial advances (Raper reference). (3) Physical considerations show that external forcing, not internal variability, played the dominant role in the transition from the "Medieval Warm Period" to "Little Ice Age" (these terms are used loosely and are, in fact, ill advised in the context of hemispheric or global temperature changes -see e.g. Bradley and Jones, 1993; Hughes and Diaz, 1994). One of the major points of Broecker's argument is that changes in the thermohaline circulation are a primary driver of climate change on this time scale. These results do not consider recent modeling studies (Free, Crowley) that demonstrate at a high significance level (>99%) that about 50% of the pre-anthropogenic (pre-1850) variance can be explained by changes in volcanism and low frequency solar irradiance. Although the latter term is still not well constrained from observational studies, there are a number of independent lines of evidence suggesting such changes (Hoyt, Lean, Lockwood). (4) It is not justifiable to argue that changes in the thermohaline circulation cause significant hemispheric or global changes in temperature. Although changes in the conveyor play a major role in the Atlantic Basin, to a first approximation changes in ocean circulation simply redistribute heat on the planet without significantly raising global temperature, or even hemispheric temperature. This conclusion is born out by very low correlations between warmth in the Greenland sector and the hemispheric indices over the last 1000 years (Crowley footnote ref.), a low correlation that is shared by coupled model experiments (Delworth citation)? In fact, sediment core data from the subtropical North Atlantic often cited as indicative of a distinct "Medieval Warm Period" and "Little Ice Age" (Keigwin Sargasso Sea), has recently been shown to be more consistent with changes in the North Atlantic Oscillation (Keigwin and Pickart), implying a zero sum pattern of regionally alternating warm and cold superimposed on far more modest hemispheric variations over the past 1000 years. This pattern itself may be forced, rather than internal in nature, and would explain the limited evidence for more dramatic cold and warm periods in regions such as Europe (see Mann, Sci Perspective, 2000). The above arguments lead us to conclude that, although the conveyor may be changing, radiative forcing perturbations were primarily responsible for centennial-millennial changes in the last 1000 years, with attendant implications for interpretation of earlier Holocene oscillations (e.g, Denton and Karlen). Furthermore, the weight of evidence indicates that the late 20th century hemispheric warming is significantly greater than the Middle Ages. Michael E. Mann Thomas J. Crowley WHO ELSE??? _______________________________________________________________________ Professor Michael E. Mann Department of Environmental Sciences, Clark Hall University of Virginia Charlottesville, VA 22903 _______________________________________________________________________ e-mail: mann@virginia.edu Phone: (804) 924-7770 FAX: (804) 982-2137 [1]http://www.evsc.virginia.edu/faculty/people/mann.html Attachment Converted: "c:\eudora\attach\broecker.htm"