Houston, we have a problem.
Admittedly, not a huge problem and not one that most people, or even most climatologists, are particularly fascinated by, but one which threads together many topics (climate models, tree rings, paleo-climate) which have been highlighted here in the past. The problem is that we have good evidence in the ice core records for very large tropical eruptions over the last 1000 years – in particular the eruptions in 1258/1259, 1452 and 1809 to 1815 – but for which many paleo-reconstructions barely show a blip in temperature. Models, in attempting to simulate this period, show varied but generally larger (and sometimes much larger) responses. The differences are significant enough to have prompted a few people to try and look into why this mismatch is occurring.
Whenever there is a mismatch between model and observation, there are, roughly speaking, at least three (non-exclusive) possibilities: the model is wrong, the observational data are wrong or the comparison is not like-with-like. There have been many examples of resolved mismatches in each category so all possibilities need to be looked at.
As described in a previous post earlier this year, Mann et al., 2012 (pdf), postulated that for extreme volcanoes, the cooling would be sufficient to saturate the growth response, and that some trees might `skip´ a ring for that year leading to a slight slippage in tree-ring dating, a potential smearing of the composite chronologies, and a further underestimate of the cooling in tree-ring based large-scale reconstructions.
This hypothesis has now been challenged by a group of authors in a comment (Anchukaitis et al.) (pdf, SI, code), who focus on the appropriateness of the tree ring growth model and the spatial pattern the volcanic climate responses. The Mann et al. response (pdf, SI) presents some further modeling and 19th century observational data in support of the original hypothesis.
Of course, there are still two other possibilities to consider. First, the models may have an excessive response. This could be due to either models responding excessively to the correct forcing, or could be related to an excessive forcing itself. There are indeed some important uncertainties in estimating the history of volcanic forcing – which involves inferring a stratospheric aerosol load (and effective radius of the particles and their distribution) from a network of sulphate peaks in ice cores in Greenland and Antarctica. For example, the forcing for the big eruption around 1453 differs by a factor of 2 in the inferred forcing (-12 W/m2 and -5.4 W/m2) in the two estimates proposed for the recent model-intercomparison (Schmidt et al., 2012). Note too that the details of how aerosols are implemented in any specific model can also make a difference to the forcing, and there are many (as yet untested) assumptions built into the forcing reconstructions.
It is also conceivable that climate models overreact to volcanic forcing – however, excellent matches to the Pinatubo response in temperature, radiative anomalies, water vapour and dynamic responses, where we know the volcanic aerosol load well, make that tricky to support (Hansen et al, 2007) (pdf, SI). (As an aside, the suggestion in this paper that the response to Krakatoa (1883) was underpredicted by the historical SST fields was partially vindicated by the results from HadSST3 which showed substantially more cooling).
The third possibility is that some tree-ring reconstructions can’t be easily compared to simple temperature averages from the models. As both the original paper and the comment suggests, there are important effects from memory from previous years in ring widths and, potentially, increases in diffuse light post-eruption promoting growth spurts. This needs to be assessed using more sophisticated forward models for tree ring growth applied to the models’ output – a feature in both the Mann et al, and Anchukaitis et al. approaches. More work is likely needed on this, and using the wider variety of model experiments coming out of CMIP5/PMIP3.
There is clearly potential for these competing hypotheses to get sorted out. Information from newly-digitised old instrumental records in the early 19th Century such as shipping records for the East India Company (Brohan et al, 2012), doesn’t support the largest modelled responses to Tambora (1815), but does suggest a response larger and more defined than that seen in some reconstructions. However, other 19th Century temperature compilations such Berkeley Earth show larger responses to Tambora – though there are spatial sampling issues there as well. There is also the potential for non-tree ring based reconstructions to provide independent confirmation of the magnitude of the response.
So while neither of the latest comments and responses provide a definitive answer to the principal problem, there is certainly lots of scope for extended and (hopefully) productive discussions.
- M.E. Mann, J.D. Fuentes, and S. Rutherford, "Underestimation of volcanic cooling in tree-ring-based reconstructions of hemispheric temperatures", Nature Geoscience, vol. 5, pp. 202-205, 2012. http://dx.doi.org/10.1038/ngeo1394
- K.J. Anchukaitis, P. Breitenmoser, K.R. Briffa, A. Buchwal, U. Büntgen, E.R. Cook, R.D. D'Arrigo, J. Esper, M.N. Evans, D. Frank, H. Grudd, B.E. Gunnarson, M.K. Hughes, A.V. Kirdyanov, C. Körner, P.J. Krusic, B. Luckman, T.M. Melvin, M.W. Salzer, A.V. Shashkin, C. Timmreck, E.A. Vaganov, and R.J.S. Wilson, "Tree rings and volcanic cooling", Nature Geoscience, vol. 5, pp. 836-837, 2012. http://dx.doi.org/10.1038/ngeo1645
- M.E. Mann, J.D. Fuentes, and S. Rutherford, "Reply to 'Tree rings and volcanic cooling'", Nature Geoscience, vol. 5, pp. 837-838, 2012. http://dx.doi.org/10.1038/ngeo1646
- G.A. Schmidt, J.H. Jungclaus, C.M. Ammann, E. Bard, P. Braconnot, T.J. Crowley, G. Delaygue, F. Joos, N.A. Krivova, R. Muscheler, B.L. Otto-Bliesner, J. Pongratz, D.T. Shindell, S.K. Solanki, F. Steinhilber, and L.E.A. Vieira, "Climate forcing reconstructions for use in PMIP simulations of the Last Millennium (v1.1)", Geoscientific Model Development, vol. 5, pp. 185-191, 2012. http://dx.doi.org/10.5194/gmd-5-185-2012
- J. Hansen, M. Sato, R. Ruedy, P. Kharecha, A. Lacis, R. Miller, L. Nazarenko, K. Lo, G.A. Schmidt, G. Russell, I. Aleinov, S. Bauer, E. Baum, B. Cairns, V. Canuto, M. Chandler, Y. Cheng, A. Cohen, A. Del Genio, G. Faluvegi, E. Fleming, A. Friend, T. Hall, C. Jackman, J. Jonas, M. Kelley, N.Y. Kiang, D. Koch, G. Labow, J. Lerner, S. Menon, T. Novakov, V. Oinas, J. Perlwitz, J. Perlwitz, D. Rind, A. Romanou, R. Schmunk, D. Shindell, P. Stone, S. Sun, D. Streets, N. Tausnev, D. Thresher, N. Unger, M. Yao, and S. Zhang, "Climate simulations for 1880–2003 with GISS modelE", Clim Dyn, vol. 29, pp. 661-696, 2007. http://dx.doi.org/10.1007/s00382-007-0255-8
- P. Brohan, R. Allan, E. Freeman, D. Wheeler, C. Wilkinson, and F. Williamson, "Constraining the temperature history of the past millennium using early instrumental observations", Climate of the Past, vol. 8, pp. 1551-1563, 2012. http://dx.doi.org/10.5194/cp-8-1551-2012