Climate Audit

Otto-Bliesner asked me how I would do a reconstruction. As I’ve said on other occasions, I said that I didn’t know. I’m really reluctant to just apple-pick some series but it’s prbably worthwhile showing that you can pick apples as well as cherries. More constructively, I think that there are some approaches that look better than others. I don’t think that Mannian data mining or forms of cherry picking are very promising. There are many problems with tree ring "site chronologies", but that doesn’t mean that Moberg’s little collection of 11 series is any magic bullet. Out of the various individual proxy studies that I’ve read, I think that two of the most interesting are Naurzbaev et al [2004], about which I reported here and a new one, Millar et al [2006], online here (I’d previously commented on a poster on the same topic here) . The topic has particular resonance for readers of this site as Millar et al. have done a detailed analysis of high-altitude (3000+ m) in the Sierra Nevadas, near the foxtails and bristlecones. Here is an excerpt from the Abstract:

Deadwood tree stems scattered above treeline on tephra-covered slopes of Whitewing Mtn (3051 m) and San Joaquin Ridge (3122 m) show evidence of being killed in an eruption from adjacent Glass Creek Vent, Inyo Craters. Using tree-ring methods, we dated deadwood to 815-1350 CE, and infer from death dates that the eruption occurred in late summer 1350 CE….Using contemporary distributions of the species, we modeled paleoclimate during the time of sympatry [the MWP] to be significantly warmer (+3.2 deg C annual minimum temperature) and slightly drier (-24 mm annual precipitation) than present,

The value of 3.2 deg C is a big and almost inconceivable number. But compare this to the following comment in Naurzabaev et al [2004] about Siberia, which I’d referred to previously, where a similar number is implied:

Trees that lived at the upper (elevational) tree limit during the so-called Medieval Warm Epoch (from A.D. 900 to 1200) show annual and summer temperature warmer by 1.5 and 2.3 deg C, respectively, approximately one standard deviation of modern temperature. Note that these trees grew 150-200 m higher (1-1.28C cooler) than those at low elevation but the same latitude, implying that this may be an underestimate of the actual temperature difference.

In my post about Naurzbaev et al., I pointed out that, when the two effects were added together, the annual and summer differences were even larger – and, as it turns out, remarkably similar to that reported by Millar et al [2006] in California.

surely it’s more reasonable to allow for the altitude effect: this would yield an estimate that the MWP annual temperature was 2.58-2.86 deg C warmer than at present and the summer MWP temperature was 3.38-3.66 deg C warmer than at present.

Millar et al [2006] applied ecological niche modeling to subfossil medieval trees located at high altitudes; ecological niche modelling is David Stockwell’s specialty so it will be interesting to see what he thinks. The locations – Whitewing Mtn and San Joaquin Ridge – are both in Mono County CA, very close to the bristlecones and foxtails. Other comments from the running text of Millar et al [2006] and I recommend reading it all:

Late Holocene climates have been described from diverse records in the Sierra Nevada and Great Basin (Woolfenden, 1996; Stine 1994). The last millennium began with a 450-year phase that corresponds to the widespread Medieval Climate Anomaly (MCA), and extended from ca. 900-1350 CE in the Sierra Nevada. Proxy records indicate this to have been a dry and warm period, where lake and river levels declined (Yuan et al., 2004; Meko et al., 2001; Stine, 1990, 1994), treelines increased (Graumlich, 1993; Graumlich and Lloyd, 1996), and glaciers retreated (Konrad and Clark, 1998). The MCA was followed by a cool phase coinciding with the northern hemispheric Little Ice Age (LIA), which extended in the Sierra Nevada from 1400-1900 CE (Clark and Gillespie, 1997). Closed lake levels remained moderately low, suggesting decreases in effective precipitation and/or runoff relative to present (Stine, 1990). Treeline elevations declined (Graumlich and Lloyd, 1996; Lloyd and Graumlich, 1997) and the largest glacial advances since the Pleistocene are recorded (Clark and Gillespie, 1997). The LIA ended ~1900 CE; early 20th century proxies record rising temperatures, precipitation increases (Graumlich, 1993) and increasing lake and river levels (Stine, 1990, 1994).
…
We modeled paleoclimates based on ecological niche theory. The six species that we identified from Whitewing Mtn do not occur together at present. We reasoned that during the period of sympatry on Whitewing Mtn, the climate must have been compatible for all the species, i.e., fundamental niche spaces overlapped (Jackson and Overpeck, 2000). Thus, conditions represented by the intersection of individual climate spaces of the deadwood species would estimate a potential climate of Whitewing Mtn during the time of the summit assemblage. A similar rationale was used by Arundel (2005).Extensive drought during the Medieval period has been further interpreted from lake sediments (Yuan et al., 2004; Benson et al., 2002; Li et al., 2000; Kleepe, in press), tree-ring reconstructions (Meko, et al., 2001), and glacial records (Konrad and Clark, 1998). Tree-ring reconstructions indicate increased temperature relative to present (Graumlich, 1993; Scuderi, 1993) and higher treelines (Graumlich and Lloyd, 1996; Lloyd and Graumlich, 1997), and pollen reconstructions show greater abundance of fir in high-elevation communities than at present (Anderson, 1990)…

The ecologic patterns and climatic estimates at Whitewing and San Joaquin Ridge corroborate studies showing significant Medieval warmth in the California region but provide evidence for differences between high and low elevations in moisture availability. Whereas mid-low elevations in the Sierra Nevada experienced extreme Medieval drought, precipitation at Whitewing appears to have been adequate to support mesic-adapted species. A projection of warm Medieval temperatures with only small decreases in precipitation at high elevations is not inconsistent with extreme drought at lower elevations indicated by other studies.

So the "regional" MWP phenomenon supposedly localized to northern Europe, is observable in Siberia and California.Readers of this site are familiar with the "site chronologies" for bristlecones and foxtails which have hockey stick shapes, which are "active ingredients" of the MBH hockey stick and other multiproxy studies.

The advantage of the approaches taken by Naurzbaev and Millar is that they are not based on correlations or on picking, but on ecological niches. While there may some hair on this concept when examined closely, intuitively it has far more appeal than (say) Thompson’s attempt to defy the “amount effect” through weak correlation arguments. The Naurzbaev and Millar studies are also very hard to reconcile with a view that certain key Hockey Team series are accurately reflecting local climate. The bristlecone and foxtail "site chronologies" used by the Hockey Team (not just in MBH but serially) show a "cold" MWP. How can these "site chronologies" be reconciled with the detailed findings of Millar et al [2006]? This is distinct from issues of CO2 fertilization. Similarly, how can the findings of Naurzbaev et al [2004] be reconciled with the Yamal site chronology which likewise argues for a "cold" MWP in this region relative to modern times (in this case, not even being consistent with the updated Polar Urals chronology.)

Moberg attempted to get away from tree ring site chronologies, but simply picked an odd selection of sediments, with his results dominated by a few questionable series. I don’t think that Moberg’s a magic bullet.

Before there was a Hockey Team, paleoclimatologists on a millennial scale studied "low frequency" vegetation changes. Millar et al [2006] is a sophisticated effort in the spirit of the paleoclimatologists before the Hockey Team. My guess is that this is the way forward – abandon Mannian attempts to extract "signals" from cherry-picked and noisy proxy series and try to develop some really good information, one site at a time.

References:
Sierra Nevada Research Center websitehttp://www.fs.fed.us/psw/programs/snrc/climate_landscape/high_elevation_sub1/medieval.shtml
Constance Millar website http://www.fs.fed.us/psw/programs/snrc/staff/millar/
Millar, C.I., J.C. King, R.D. Westfall, H.A. Alden, and D.L. Delany. 2006. Late Holocene forest dynamics, volcanism, and climate change at Whitewing Mountain and San Joaquin Ridge, Mono County, Sierra Nevada, CA, USA. Quaternary Research. In Press. (Article in PDF Format; Figures in PDF Format) http://www.fs.fed.us/psw/programs/snrc/staff/millar/Whitewing_txt.pdf