Guest commentary from Tim Osborn, Tom Melvin and Keith Briffa, Climatic Research Unit, UEA

Records of tree-ring characteristics such as their width (TRW) and density (usually the maximum density of the wood formed towards the end of the growing season – the “maximum latewood density” – MXD) are widely used to infer past variations in climate over recent centuries and even millennia. Chronologies developed from sites near to the elevational or latitudinal tree lines often show sensitivity to summer temperature and, because of their annual resolution, absolute dating and relatively widespread nature, they have contributed to many local, continental and hemispheric temperature reconstructions. However, tree growth is a complex biological process that is subject to a range of changing environmental influences, not just summer temperature, and so replication, coherence and consistency across records and other proxies are an important check on the results.

Tree-ring records have greater replication (both within a site and between nearby sites) than other types of climate proxy. Good replication helps to minimise the influence of random localised factors when extracting the common signal, and it also allows the comparison of information obtained from different independent sets or sub-sets of data. If independent sets of data – perhaps trees with different mean growth rates or from different sites – show similar variations, then we can have greater confidence that those variations are linked to real variations in climate.

In a new QSR paper (Briffa et al., 2013), (BEA13) we have used these approaches to re-assess the combined tree-ring evidence from the Yamal and Polar Urals region (Yamalia) of northern Siberia, considering the common signal in tree-growth changes at different sites and in subsets of data defined in other ways. Together with our Russian colleagues and co-authors, we have incorporated many new tree-ring data, to increase the replication and to update the chronology to 2005 and have reassessed the inferences about summer temperature change that can be drawn from these data. The paper is published as an open-access paper (no paywall) and supplementary information including the raw tree-ring and instrumental temperature data are available from our website.

Figure 1 illustrates our inferences about past summer temperature variations. Low tree-growth periods for which the inferred summer temperatures are approximately 2.5°C below the 1961-90 reference are apparent in the 15-year smoothed reconstructions (Figure 1d), centred around 1005, 1300 (Figure 1b), 1455 (Figure 1c), 1530, particularly the 1810s where the inferred cooling reaches -4 or even -6°C for individual years (Figure 1a), and the 1880s. These temperature estimates will be interesting for the current debate about the representation of volcanically-induced cooling in temperature reconstructions, and for testing of climate model simulations.

There are numerous periods (Figure 1d) of one or two decades with relatively high growth (and inferred summer temperatures close to the 1961-90 level) but at longer timescales (Figures 1e and 1f) only the 40-year period centred at 250 CE appears comparable with 20th century warmth. This early warm period was both preceded and followed by periods of low ring width and so the central estimates of the temperature reconstruction averaged over the warmest 100-year period near the 3rd century CE (205-304 CE) are 0.4°C cooler than the 1906-2005 mean. Allowing for chronology and reconstruction uncertainty, we find that the mean of the last 100 years of the reconstruction is likely warmer than any century in the last 2000 years in this region.

Figure 1 (from Fig. 13 of BEA13). Summer temperature reconstructions based on either the Yamal ring-width chronology (red line, orange confidence intervals) or by combining information from the Yamal and Polar Urals ring-width chronologies and the Polar Urals density chronology (blue line, blue confidence intervals). The latter is shorter because the Polar Urals data are shorter and also has two versions that differ in how they are calibrated and in the summer temperature that they represent (in panels (a)-(e) it represents mean June–August temperature shown by the black dotted lines, while in panel (f) it represents mean June–July temperature shown by black continuous lines). Each panel shows a different time period and degree of smoothing; the values near to the end of the smoothed series are more uncertain than shown here due to the presence of end effects on the spline filters. The low-frequency agreement between the series is expected because the Yamal ring-width data are common to both reconstructions.

A response to the critics

The publication of our paper provides a timely opportunity to revisit and respond to a series of unfounded criticisms that have been levelled at our work in recent years, mostly originating from Steve McIntyre at the ClimateAudit blog, though they have been widely repeated and embellished by other commentators.

It is of course usual for results to be improved and superseded as science progresses. Our new Yamalia ring-width chronology differs from the Yamal chronology published by Briffa (2000) – see Figure 2a for a comparison. The very recent values are now lower (and extend by a decade more), but so are the estimates around 1000 CE. The consequent differential between medieval and modern growth is hardly changed. The period of high growth centred near to 250 CE (noted above) is also relatively unchanged, and is now the most prominent pre-20th century period of anomalous growth in the last 2000 years. These changes are because of genuine scientific progress, not because – as our critics have claimed – we had previously presented a deceptive chronology. They arise from extra data collection and, particularly, developments in tree-ring standardization methods (see the paper for details).

Figure 2. (a) Comparison of the Briffa (2000) Yamal ring-width chronology (red) and the new Yamalia ring-width chronology (black). (b) Comparison of the new Yamalia ring-width chronology (black) and two chronologies that have been promoted by critics of our work, but which turn out to be biased: the Polar Urals “update” chronology (purple; from Esper et al., 2002) and the Yamal chronology with modern data coming only from the Khadyta River site (blue). All series were scaled to have unit variance before being smoothed with a 10-year filter.

Figure 2b compares the new Yamalia chronology with two alternative chronologies heavily promoted by McIntyre and others – the so-called Polar Urals “update” chronology and a Yamal chronology using modern samples from the Khadyta River site. Both chronologies present a different picture of the difference between peak medieval and peak modern growth rates, with elevated growth around 1000 CE and suppressed growth in the 20th century. Our paper demonstrates that these two alternative chronologies are flawed.

The real Yamal deception

Some background is perhaps needed regarding our preferred chronologies. Briffa et al. (1995) developed chronologies from Polar Urals ring width and density data. Subsequently, Briffa (2000) presented a 2000-year ring width chronology from nearby Yamal, which had much better replication (more trees) than the Polar Urals data and was therefore preferred. The Polar Urals data were later supplemented by additional samples which were used by Esper et al. (2002). Even including these additional samples the Yamal chronology remained better replicated: of the 1213 overlap years, the Briffa (2000) Yamal has 4 years with samples from less than 10 trees, while the “updated” Polar Urals chronology has 264 years with data from less than 10 trees, many of them in the medieval period (see here for more details). The additional sub-fossil data used in our new paper further increases the replication of the Yamal chronology compared with the Polar Urals chronology (Figure EC1 in the SI of the new paper). On the basis of replication and the strength of the common signal, the Yamal record was, and remains, superior to the Polar Urals chronology.

1: Why we didn’t use the Polar Urals “update”

We have been criticised for not archiving the Polar Urals “update” data. The “update” data were in fact archived at the ITRDB thirteen years ago. We have been criticised for not publishing an updated Polar Urals chronology using the updated data (and accused of worse here). The supposed reason for our decision not to do this was that the ‘update’ does not support our supposedly desired message of unprecedented modern warmth (because they appear to suggest that tree growth rate was greater during earlier times including the medieval period; Figure 2b, compare purple and black lines).

However, as reported in BEA13, it turns out that during the medieval period these Polar Urals “updates” are dominated by samples taken from the root collars of trees. Ring widths measured in such root-collar samples tend to be systematically larger than equivalent rings measured higher in the boles (stems) of the same trees. The reason for larger tree-ring widths during medieval times in the Polar Urals “updates” is now clear: it is because more samples were from the root collar with their inherently wider rings. Interpreting this as evidence for warmer temperatures is wrong.

Conclusion: the so-called “Polar Urals update” chronology is severely biased and should not be used as evidence of past changes in temperature; nor should our critics present it as evidence that we had committed scientific fraud by failing to publish a chronology using these data.

2: The Yamal record was not biased by omitting data

CRU has been accused of deception by presenting a Yamal tree-ring chronology biased by the omission of otherwise suitable data. A particular theme, originating again from ClimateAudit, is that tree-ring data from Khadyta River had not been used and would have dramatically altered the character of the chronology – and the NH temperature reconstructions that used the Yamal chronology – if these data had been used (Figure 2b, compare blue and black lines).

As reported in BEA13, through collaboration with our Russian colleagues who have extensive knowledge of tree-rings in this region, we have learnt that the Khadyta River site has problems related to the particular site conditions that differ from other sites in this region, and maybe influenced by changing permafrost. Certainly the trees have reduced growth and appear to be unhealthy, and some even dying. Thus the Khadyta River data that some claimed as being more representative than the data we used turn out to have a common signal that is inconsistent with the majority of site chronologies in this region. They could potentially bias the Yamal chronology had they been included and so for this reason we excluded these data from the main analysis in the new paper.

Conclusion: claims of a deceptive and biased Yamal chronology turn out to rely on outlier data that should be omitted; our new research, based on a greatly expanded dataset, supports the finding that tree-growth (and inferred summer temperature) in this region are likely greater in the last 100 years than for any previous century in the last 2000 years.

3: We did not withhold a combined Yamal and Polar Urals chronology

Separately, some of our incomplete and unpublished work on the Yamal and Polar Urals tree-ring data has been the subject of multiple requests under UK FOI/EIR legislation. (See this previous post for background). To be clear, this was not a request for the raw data that we were using in this area of northern Russia – the raw data were and are freely available. Instead, the request was for a tree-ring chronology that formed part of work that was, at the time, still ongoing.

The EIR has a (very sensible) exemption for material which is unfinished, incomplete or still in the course of completion. Our university (UEA) therefore refused the requests to release our incomplete research (see responses here and here). Steve McIntyre appealed and UEA reconsidered the issues but upheld the original decision. McIntyre then complained to the Information Commissioner’s Office (ICO). The ICO upheld UEA’s decision and rejected McIntyre’s complaint. McIntyre then appealed to the First-Tier Information Tribunal. Two weeks ago, after more than two years defending our right to publish our research at a time when we considered it to be complete rather than at a time dictated to us by Steve McIntyre, the Information Tribunal finally dismissed McIntyre’s appeal.

The research that was the subject of this information request has now been – as we said all along that it would be – completed and published, coincidentally, within days of the Information Tribunal’s decision. Our publication of this work contradicts McIntyre’s explicit accusations that we were hiding the requested chronology because it would have exposed long-standing scientific fraud on our part. These accusations were, and remain, baseless and mistaken.

Over the years, McIntyre has advanced a number of other criticisms of our tree-ring work in northwestern Eurasia. We note here that these too are also wrong.: 1) the original Polar Urals chronology was not wrongly cross-dated as claimed in a 2005 submission to Nature by McIntyre and McKitrick. When we demonstrated this in our response, Nature decided to publish neither their comment nor our response. It is worth noting that this rejection, nor any acknowledgement of his erroneous conclusions, were ever mentioned by McIntyre on his blog. (2) The Grudd (2008) Tornetrask chronology, promoted by some because of its elevated medieval growth (and implied much greater warmth) relative to the modern period, is biased by the issues noted in Melvin et al. (2013).

In conclusion, criticisms of our work have been based on misconceptions and misinformation. The so-called Polar Urals “update” chronology promoted by our critics turns out to be biased by inclusion of samples from tree root collars. The Khadyta River tree-ring data, whose exclusion from the Yamal chronology was portrayed as a severe example of cherry-picking to obtain a pre-conceived outcome, are from trees that appear to be dying and do not have a common signal with other regions. An updated Tornetrask chronology, with apparently elevated medieval warmth, turns out to be biased by combining incompatible groups of measurements.

That the critics have promoted a series of results that have turned out to be flawed is unfortunate but not in itself reason to complain – as science progresses it is usual for results to be improved and superseded. What can be condemned, however, is the long campaign of allegations of dishonesty and scientific fraud made against us on the basis of these false claims. That is the most disquieting legacy of Steve McIntyre and ClimateAudit. The real Yamal deception is their attempt to damage public confidence in science by making speculative and scandalous claims about the actions and motivations of scientists while cloaking them in a pretense of advancing scientific knowledge.

Links to other relevant information

CRU response to Yamal criticisms in 2009

CRU comments on the Ross McKitrick 2009 article published in the Financial Post (17 June 2010).

Response to comments about the dating of the Polar Urals chronology, the reliability of the early parts of the Polar Urals reconstruction, and our view that the Yamal chronology was more reliable.

References

1. K.R. Briffa, T.M. Melvin, T.J. Osborn, R.M. Hantemirov, A.V. Kirdyanov, V.S. Mazepa, S.G. Shiyatov, and J. Esper, "Reassessing the evidence for tree-growth and inferred temperature change during the Common Era in Yamalia, northwest Siberia", Quaternary Science Reviews, vol. 72, pp. 83-107, 2013. http://dx.doi.org/10.1016/j.quascirev.2013.04.008
2. K.R. Briffa, "Annual climate variability in the Holocene: interpreting the message of ancient trees", Quaternary Science Reviews, vol. 19, pp. 87-105, 2000. http://dx.doi.org/10.1016/S0277-3791(99)00056-6
3. K.R. Briffa, P.D. Jones, F. Schweingruber, S. Shiyatov, and E. Cook, "Unusual twentieth-century summer warmth in a 1,000-year temperature record from Siberia", Nature, vol. 376, pp. 156-159, 1995. http://dx.doi.org/10.1038/376156a0
4. J. Esper, "Low-Frequency Signals in Long Tree-Ring Chronologies for Reconstructing Past Temperature Variability", Science, vol. 295, pp. 2250-2253, 2002. http://dx.doi.org/10.1126/science.1066208
5. H. Grudd, "Torneträsk tree-ring width and density ad 500–2004: a test of climatic sensitivity and a new 1500-year reconstruction of north Fennoscandian summers", Climate Dynamics, vol. 31, pp. 843-857, 2008. http://dx.doi.org/10.1007/s00382-007-0358-2
6. T.M. Melvin, H. Grudd, and K.R. Briffa, "Potential bias in 'updating' tree-ring chronologies using regional curve standardisation: Re-processing 1500 years of Tornetrask density and ring-width data", The Holocene, vol. 23, pp. 364-373, 2013. http://dx.doi.org/10.1177/0959683612460791

June’s open thread…

Guest post from PubPeer.com

The process of reviewing published science is constantly occurring and is now commonly being called post-publication peer review. It occurs in many places including on blogs such as this one, review articles, at conferences around the world, and has even been encouraged on the websites of some journals. However, the process of recording and searching these comments is, unfortunately, inefficient and underused by the larger scientific community for several reasons: To successfully impact the publication process, this database of knowledge has to accomplish two important tasks. First it requires participation by a large part of a given scientific community so that it reflects an average impression instead of an outlier’s impression. Second, it requires that the collective knowledge is centralized and easy to search in order find out what the community collectively thinks about an individual paper or a body of work. A recent initiative, the San Francisco Declaration on Research Assessment (DORA), echoes many of these same concerns.

In an attempt to assemble such a database, a team of scientists, have put together a website called PubPeer.com that is searchable and encourages participation by the larger scientific community. With a critical mass of usage an organized system of post-publication review could improve both the process of scientific publication as well as the research that underlies those publications.

Those of us involved in the creation of PubPeer.com believe that in an ideal world, a scientist’s main goal would be to discover something interesting about the world and simply report it to other scientists to use and build upon. This idealistic view of the scientific process is however not matched in reality because, for academic scientists, our publications count for much more than a simple contribution to the scientific record. For example, the majority of candidates are eliminated from consideration for tenure track positions at a major universities based on the names of the journals that have published their recent findings.

Review committees use this method because publications are the best measure of past and potential scientific output, but by potentially overvaluing “high impact” journal names, these committees and study sections effectively defer to journal editors to help them choose the best candidates for jobs and grants. However, these journals select their articles based on more than just good science – the papers also need to be of ‘wider interest’ and this can sometimes skew the publications towards ‘exciting’ results over those that are more measured, and perhaps more likely to be correct (for instance). The sometimes disproportinate attention given to a high profile paper also makes it a tempting target for more unscrupilous scientists.

It’s never going to be possible for us to thoroughly read all of the papers submitted to a job advertisement, nor all of the papers referenced in grant applications, but we can easily reduce the importance that journal names play in decisions and replace it with something that is more meaningful and directly in our hands instead of the hands of publishers. After reading any publication, we all have impressions about whether the reported observations are useful, interesting, elegant, irrelevant, flawed, etc. If a particular scientific field that is interested in a given publication were able to compile all of it’s impressions of that publication, that collective information would be infinitely more useful to search committees and study sections than the name of the journal in which it was published.

Outlined below are a few aspects of PubPeer.com that differentiate it from the current post-publication review systems and which will hopefully make it more widely used.

• A key issue that we have decided on is the importance of anonymity. One of the reasons that we have never commented on articles directly on journal websites is because the colleagues whose publications we are most qualified to comment on are likely reviewing our publications and grant proposals. Even the most well-intentioned criticism could potentially irk these potential reviewers. Since publications are so precious to everyone’s future career advancement, there is a huge psychological barrier for early stage scientists to attach names to any comments that could be considered critical.

  Therefore, in order to encourage as much participation in this post-publication review process as possible, PubPeer allows comments to be left anonymously if someone is so inclined. Critics of this feature sometimes email us to point out that anonymity allows for baseless slander or to proclaim that a commenter’s name is essential for judging the validity of a comment. We strongly disagree with this second point because good comments are good regardless of whether they come from a senior scientist or a graduate student. We can all judge for ourselves the content of comments and on PubPeer it is possible to vote the good comments up and the bad comments down into the noise so that community as a whole can decide together what is worth paying attention to. Baseless defamation, rumors, and ad hominen attacks are not tolerated at all and are immediately removed from the site.

  The people involved with PubPeer are all active scientists and we are trying to remain anonymous for the time being for several reasons: 1) we can imagine scenarios in which pressure could be put on us to remove/alter comments if our identities were known and 2) we would like to protect our families and private bank accounts from the more litigious among our readers.

• A main drawback of the current practice of post-publication peer review is that the reviews can be spread across many different blogs and journal websites. If one wants to know what the community thinks of a given body of work (whether it be a discipline, an author’s output, a university department, etc.) it takes a major time investment to track down the information from all of these different sources. PubPeer provides a centralized and easily searchable database that contains comments on all published articles.
• PubPeer also provides for a system of alerts. In order to be effective, authors and others interested should be able to be alerted to comments on their favorite publications or topics. Pubpeer automatically notifies corresponding authors of new comments on their articles and anyone can set up email alerts on articles they find interesting.

We’d like to thank realclimate.org for this invitation to explain PubPeer and we welcome any suggestions, comments, criticism, or questions on the contact page pubpeer.com/contact or in the comments section below. Many conversations have already started on the site both with and without author responses that are sometimes quite detailed (a list of all of the most recent articles receiving comments can be found at pubpeer.com/recent). Some comments have already led to important corrections in high profile studies (see also Nature News, Science Insider, http://www.realclimate.org/?p=15488

by Jill and David Archer

This month’s open thread.

There has been an unusual surge of interest in the climate sensitivity based on the last decade’s worth of temperature measurements, and a lengthy story in the Economist tries to argue that the climate sensitivity may be lower than previously estimated. I think its conclusion is somewhat misguided because it missed some important pieces of information (also see skepticalscience’s take on this story here).

The ocean heat content and the global mean sea level height have marched on.

While the Economist referred to some unpublished work, it missed a new paper by Balmaseda et al. (2013) which provides a more in-depth insight. Balmaseda et al suggest that the recent years may not have much effect on the climate sensitivity after all, and according to their analysis, it is the winds blowing over the oceans that may be responsible for the ‘slow-down’ presented in the Economist.

It is well-known that changes in temperature on decadal time scales are strongly influenced by natural and internal variations, and should not be confused with a long-term trend (Easterling and Wehner, 2009;Foster and Rahmstorf, 2011).

An intensification of the trades has affected surface ocean currents called the subtropical gyres, and these changes have resulted in a predominance of the La Nina state. The La Nina phase is associated with a lower global mean temperature than usual.

Balmaseda et al’s results also suggested that a negative phase of the pacific decadal oscillation (PDO) may have made an imprint on the most recent years. In addition, they found that the deep ocean has warmed over the recent years, while the upper 300m of the oceans have ‘stabilised’.

The oceans can be compared to a battery that needs to be recharged after going flat. After the powerful 1997-98 El Nino, heat flowed out of the tropical oceans in order to heat the atmosphere (evaporative cooling) and the higher latitudes. The warming resumed after the ‘deflation’, but something happened after 1998: since then, the warming has involved the deep ocean to a much greater extent. A weakening of the Atlantic meridional overturning circulation (MOC) may have played a role in the deep ocean warming.

The recent changes in these decade-scale variations appear to have masked the real accumulation of heat on Earth.

The new knowledge from this paper, the way I read it, is the revelation of the role of winds for vertical mixing/diffusion of heat in a new analysis of the world oceans. Their results were derived through a set of different experiments testing the sensitivity to various assumptions and choices made for data inclusion and the ocean model assimilation set-up.

The analysis involved a brand new ocean analysis (ORAS4; Balmaseda et al., 2013) based on an optimal use of observations, data assimilation, and an ocean model forced with state-of-the-art description of the atmosphere (reanalyses).

By running a set of different experiments with the ocean model, including different conditions, such as surface winds and different types of data, they explored which influence the different conditions have on their final conclusion.

The finding that the winds play a role for the state of the warming may not be surprising to oceanographers, although it may not necessarily be the first thing a meteorologist may consider.

Other related discussions: OSS

References

1. M.A. Balmaseda, K.E. Trenberth, and E. Källén, "Distinctive climate signals in reanalysis of global ocean heat content", Geophysical Research Letters, pp. n/a-n/a, 2013. http://dx.doi.org/10.1002/grl.50382
2. D.R. Easterling, and M.F. Wehner, "Is the climate warming or cooling?", Geophysical Research Letters, vol. 36, 2009. http://dx.doi.org/10.1029/2009GL037810
3. G. Foster, and S. Rahmstorf, "Global temperature evolution 1979–2010", Environmental Research Letters, vol. 6, pp. 044022, 2011. http://dx.doi.org/10.1088/1748-9326/6/4/044022
4. M.A. Balmaseda, K. Mogensen, and A.T. Weaver, "Evaluation of the ECMWF ocean reanalysis system ORAS4", Quarterly Journal of the Royal Meteorological Society, pp. n/a-n/a, 2013. http://dx.doi.org/10.1002/qj.2063

Guest commentary by Darrell Kaufman (N. Arizona U.)

In a major step forward in proxy data synthesis, the PAst Global Changes (PAGES) 2k Consortium has just published a suite of continental scale reconstructions of temperature for the past two millennia in Nature Geoscience. More information about the study and its implications are available at the FAQ on the PAGES website and the datasets themselves are available at NOAA Paleoclimate.

The main conclusion of the study is that the most coherent feature in nearly all of the regional temperature reconstructions is a long-term cooling trend, which ended late in the 19th century, and which was followed by a warming trend in the 20th C. The 20th century in the reconstructions ranks as the warmest or nearly the warmest century in all regions except Antarctica. During the last 30-year period in the reconstructions (1971-2000 CE), the average reconstructed temperature among all of the regions was likely higher than anytime in at least ~1400 years. Interestingly, temperatures did not fluctuate uniformly among all regions at multi-decadal to centennial scales. For example, there were no globally synchronous multi-decadal warm or cold intervals that define a worldwide Medieval Warm Period or Little Ice Age. Cool 30-year periods between the years 830 and 1910 CE were particularly pronounced during times of weak solar activity and strong tropical volcanic eruptions and especially if both phenomena often occurred simultaneously.

Figure: Thirty-year mean relative temperatures for the seven PAGES 2k continental-scale regions arranged vertically from north to south.

The origin of the ‘PAGES 2k Network‘ and its activities can be found here and consists of nearly 80 individual collaborators. The Consortium’s collection of local expertise and proxy records was transformed into a synthesis by a smaller team of lead authors, but the large author list recognizes that the expertise of the wider team was essential in increasing the range of data used and interpreting it.

In addition to the background available at the FAQ, I think it is important to also highlight some aspects of the analytical procedures behind the study and the vital contributions of three young co-authors.

The benefit of the ‘regions-up’ approach embodied in the PAGES-2k consortium is that it made it easy to take advantage of local expertise and include a large amount of new data that would have been more difficult to assemble for a centralized global reconstruction. However, being decentralized, the groups in different regions opted for different methodologies for building their default reconstructions. While justifiable, this does raise a question about the impact different methodologies would have. To address this, the synthesis team (ably led by Nicholas McKay) applied three particular reconstruction methods to all of the regions, as well as looking at the basic area-averaged and weighted composites. He further analyzed the site-level records individually and without many of the assumptions that underlie the regional temperature reconstructions. These results show that the long-term cooling trend and recent warming are dominant features of the dataset however you analyze it. There is a sizable fraction of the records that do not conform to the continental averages, highlighting the spatial variability and/or the noise level in specific proxies.

One of the new procedures used to reconstruct temperature is an approach developed by Sami Hanhijärvi (U. Helsinki), which was also recently applied to the North Atlantic region. The method (PaiCo) relies on pairwise comparisons to arrive at a time series that integrates records with differing temporal resolutions and relaxes assumptions about the relation between the proxy series and temperature. Hanhijärvi applied this procedure to the proxy data from each of the continental-scale regions and found that reconstructions using different approaches are similar and generally support the primary conclusions of the study.

Regions where this study helps clarify the temperature history are mainly in the Southern Hemisphere. We include new and updated temperature reconstructions from Antarctica, Australasia and South America. The proxy records from these three regions come from many sources, ranging from glacier ice to trees and from lake sediment to corals. Raphael Neukom (Swiss Federal Research Institute WSL and University of Bern) played a key role in the analyses across the Southern Hemisphere. He used principal components regression (Australasia), a scaled composite (Antarctica), and an integration of these two approaches (South America) to create the time series of annual temperature change.

Inevitably, assembling such a large and diverse dataset involves many judgement calls. The PAGES-2k consortium has tried to assess the impact of these structural decisions by using multiple methods, but we hope that this synthesis is really just the start of a more detailed analysis of regional temperature trends and we welcome constructive suggestions for improvements.

References

1. M. Ahmed, K.J. Anchukaitis, A. Asrat, H.P. Borgaonkar, M. Braida, B.M. Buckley, U. Büntgen, B.M. Chase, D.A. Christie, E.R. Cook, M.A.J. Curran, H.F. Diaz, J. Esper, Z. Fan, N.P. Gaire, Q. Ge, J. Gergis, J.F. González-Rouco, H. Goosse, S.W. Grab, N. Graham, R. Graham, M. Grosjean, S.T. Hanhijärvi, D.S. Kaufman, T. Kiefer, K. Kimura, A.A. Korhola, P.J. Krusic, A. Lara, A. Lézine, F.C. Ljungqvist, A.M. Lorrey, J. Luterbacher, V. Masson-Delmotte, D. McCarroll, J.R. McConnell, N.P. McKay, M.S. Morales, A.D. Moy, R. Mulvaney, I.A. Mundo, T. Nakatsuka, D.J. Nash, R. Neukom, S.E. Nicholson, H. Oerter, J.G. Palmer, S.J. Phipps, M.R. Prieto, A. Rivera, M. Sano, M. Severi, T.M. Shanahan, X. Shao, F. Shi, M. Sigl, J.E. Smerdon, O.N. Solomina, E.J. Steig, B. Stenni, M. Thamban, V. Trouet, C.S. Turney, M. Umer, T. van Ommen, D. Verschuren, A.E. Viau, R. Villalba, B.M. Vinther, L. von Gunten, S. Wagner, E.R. Wahl, H. Wanner, J.P. Werner, J.W. White, K. Yasue, and E. Zorita, "Continental-scale temperature variability during the past two millennia", Nature Geoscience, vol. 6, pp. 339-346, 2013. http://dx.doi.org/10.1038/ngeo1797
2. S. Hanhijärvi, M.P. Tingley, and A. Korhola, "Pairwise comparisons to reconstruct mean temperature in the Arctic Atlantic Region over the last 2,000 years", Climate Dynamics, 2013. http://dx.doi.org/10.1007/s00382-013-1701-4

Eric Steig

It is well known that ice shelves on the Antarctic Peninsula have collapsed on several occasions in the last couple of decades, that ice shelves in West Antarctica are thinning rapidly, and that the large outlet glaciers that drain the West Antarctic ice sheet (WAIS) are accelerating. The rapid drainage of the WAIS into the ocean is a major contributor to sea level rise (around 10% of the total, at the moment).

All of these observations match the response, predicted in the late 1970s by glaciologist John Mercer, of the Antarctic to anthropogenic global warming. As such, they are frequently taken as harbingers of greater future sea level rise to come. Are they?

Two papers published this week in Nature Geoscience provide new information that helps to address this question. One of the studies (led by me) says “probably”, while another (Abram et al.) gives a more definitive “yes”.

The somewhat different details of the two papers appear to have hopelessly confused many journalists (though the Christian Science Monitor has an excellent article, despite a somewhat misleading headline), but both are really just telling different aspects of the same story.

There is already strong evidence that anthropogenic forcing has played a significant role in the collapse of ice shelves on the Antarctic Peninsula, cause by significant melting at the surface during summer. The warm summer air temperatures have been related to an increase in the “Southern Annular Mode” (SAM), essentially the strength of the circumpolar westerlies. Increased CO₂ is clearly part of the forcing of the observed positive trend in the SAM, though a larger player is likely to be ozone depletion in the stratosphere. Nevertheless, the short length of the observations – of both the ice sheet and climate – make it difficult to assess to what extent these changes are unusual. There is evidence for one ice shelf that a collapse like that observed in the 1990s has not occurred since at least the mid-Holocene, but comparable evidence is lacking elsewhere.

The connection between climate change and glacier response is more complex for the West Antarctic Ice Sheet than the Peninsula. As on the Peninsula, temperatures over the WAIS have risen significantly in the last few decades, but this is a symptom, rather than a cause. For WAIS, the culprit for the rapid thinning of ice shelves is increased delivery of warm ocean water to the base of the ice shelves. This isn’t due to a warming ocean (though the deep water off the Antarctic coast line is indeed warming), but to changes in the winds that have forced more circumpolar deep water onto the continental shelf. Circumpolar deep water, at about +2°C, is very hot compared with the in situ melting point of glacier ice. In a series of papers, we’ve shown that the warmer temperatures observed over the WAIS are the result of those same atmospheric circulation changes, which are not related to the SAM, but rather to the remote forcing from changes in the tropical Pacific: changes in the character of ENSO (Steig et al., 2012; Ding et al., 2011; 2012).

As on the Peninsula, there is evidence of anthropogenic forcing for the WAIS too: anomalous conditions since the 1980s in the tropical Pacific are characteristic of the expected fingerprint of global warming (e.g. Trenberth and Hoar, 1997; Collins et al., 2010). Still, as on the Peninsula, the short length of the instrumental observations make it difficult to say anything very definitive about long term trends.

Both our paper and that of Abram et al. add to our understanding of recent climate, glacier, and ice sheet changes in Antarctica by placing them into a longer-term context. Amidst the continuous chatter in the blogosphere about the strengths and limitations about “multiproxy” studies, these studies may be a refreshing return to simpler methods relying on just one type of “proxy”: data from ice cores. While ice core data aren’t perfect proxies of climate, they come pretty close, and aren’t subject to the same kinds of uncertainties that are unavoidable in biological proxies like tree rings.

Our study is the culmination of about a decade of ice core drilling and analysis in West Antarctica, through the ITASE program and the WAIS Divide ice core project. I’m the lead author on the paper but the author list is rightfully long; a lot of people have been involved in drilling and analyzing cores all across Antarctica.

The only “proxy” we use are oxygen isotope ratios. Oxygen isotope ratios (δ¹⁸O) in polar snow are well known to be correlated with temperature, and the underlying physics of the relationship is very well understood. In our study, we compile all the available δ¹⁸O data from high-resolution well-dated ice cores in West Antarctica and take a look at the average variability through the last 200 years. We also include data from the new WAIS Divide ice core that goes back 2000 years (actually, this core goes back to 68,000 years, and is annually resolved back to at least 30,000 years, but that’s a story for another time).

The average of the records for the last 50 years looks very much like temperature records from the last 50 years, with scaling of about 0.5‰/°C, exactly as expected, providing yet another piece of evidence that recent warming in West Antarctica has been both rapid and widespread (see the figure below). A critical point, though, is that it isn’t necessary to use the δ¹⁸O data as a proxy for temperature. Because the physics controlling δ¹⁸O is well understood, and we are able to implement δ¹⁸O in climate models, we can actually just use δ¹⁸O as a proxy for, well, δ¹⁸O. This simplifies the problem from “how significant is the recent warming?” to “how significant is the recent rise in δ¹⁸O”? We’ve shown previously, and show again in this paper, that δ¹⁸O in West Antarctic precipitation reflects the relevant changes in atmospheric circulation just as well (if not better) than temperature or other conventional climate variables do. Putting δ¹⁸O into a GCM and using the same experiments that reproduce the observed warming over West Antarctica also produces the observed δ¹⁸O increase in the last 50 years.

Figure 1. (a) Comparison of averaged δ¹⁸O (blue) across West Antarctica with the recent temperature record of Bromwich et al. (2013) from central West Antarctica (yellow). The light blue background is the decadal smoothed values +/- 1 standard error assuming Gaussian statistics. (b) Number of records used, and probability that the decadal average is as elevated as the 1990s (green).

Data sources: Most of the data for this figure have been available at http://nsidc.org/data/NSIDC-0425.html for some time. There’s a new location (which will link to the old one) where more recent data sets will be placed, but it’s not all up yet: http://nsidc.org/data/nsidc-0536.html.

Our results show that the strong trend in δ¹⁸O in West Antarctica in the last 50 years is largely driven by anomalously high δ¹⁸O in the most recent two decades, particularly in the 1990s (less so the 2000s). This is evident in the temperature data as well (top panel of the figure). The 1990s were also very anomalous in the tropics — there were several large long-lived El Niño events with a strong central tropical Pacific expression, as well as only very weak La Niña events. As in the tropics, so in West Antarctica: the 1990s were likely the most anomalous decade of the last 200 years.

Our results thus show that, indeed, recent decades in West Antarctica, which have been characterized by very rapid warming, and very rapid loss of ice from the West Antarctic Ice Sheet, are highly unusual. Nevertheless, some caution is in order in interpreting this to mean that current rates of rapid ice loss from West Antarctica represent a long term trend. What we’ve observed is unusual, but it is also dominated by decadal climate variability, and can’t be considered “unprecendeted”. Furthermore, our statistical confidence that recent decades are truly exceptional is low. Our data suggest that there is about a 30% chance the 1940s were just as anomalous as the 1990s, and the 1830s have about a 10% chance of being like the 1990s. Based on the relatively small amount of available evidence from the tropics, both the 1940s and the 1830s were similarly characterized by long-lived El Niños. Looking at the very long term record from the WAIS Divide ice core, it appears that similar conditions could have occurred about once per century over the last 2000 years. Hence our answer to the question, “are the observations of the last few decades a harbinger of continued ice sheet collapse in West Antarctica?”, is tentative: “Probably”.

Anyone expecting a more dramatic result need only turn to the other new ice core paper in Nature Geoscience. Last year, Rob Mulvaney and others from the British Antarctic Survey (BAS), along with French, American, and German colleagues, reached a very similar conclusion to ours, from an ice core from James Ross Island, on the northern Antarctic Peninsula. We discussed that paper at Realclimate last year. With δ¹⁸O data alone, it was possible to demonstrate only that recent warming on James Ross Island was “unusual”. The new paper, led by Nerelie Abram, adds a record of melt layers in the ice core to the assessment. The findings: a veritable Antarctic ice hockey stick.

Figure 2. δ¹⁸O (scaled to temperature) and melt layer frequency from the James Ross Island ice core.

Abram et al.’s paper is elegant in its simplicity. The key thing that matters to the ice shelves on the Antarctic Peninsula is how much melting occurs in summer, and this is almost exactly what Abram et al. are looking at. I say “almost” because formation of melt layers requires both that melting occurs and that it gets preserved, which depends a bit on the snow structure, the previous winter temperature, etc. But the results are unequivocal: there’s about 5 times the fraction of melt layers in the core as there has been on average over previous decades, and at least twice the maximum of any time before about the 1950s. The amount of melting occurring now is greater than at any time in the past 1000 years. If there has ever been a question about whether the “hockey stick” shape of Northern Hemisphere temperatures extends to at least some areas of the Southern Hemisphere, this record provides a decisive and positive answer.

Why the difference between the Peninsula and the WAIS? After all, both locations are warming at about the same rate. We could speculate that if there were melt layers in the WAIS cores, they would also show a significant increase like the James Ross Island core does. (It’s too cold at all the WAIS sites to have summer melting at all, so such information isn’t available.) I don’t think that is likely though. More important is the specific location of James Ross Island, on the eastern side of the Antarctic Peninsula. On the western Antarctic Peninsula, temperature trends are greatest in winter and spring, just as they are over the WAIS, and we’ve argued elsewhere that the causes are similar: changes in regional circulation forced by anomalous conditions in the tropics (Ding and Steig, in press). But it is on the eastern Peninsula that the most rapid summer warming has occurred, and where the surface-melting has caused ice shelf collapse (indeed, James Ross Island wasn’t really an island until 1995, when the Prince Gustav ice shelf collapsed). Both statistical assessments and modeling results show that the trend in the SAM accounts for this warming trend. As I noted in the introduction to this post, the SAM trend is partly explained by ozone depletion in the stratosphere, and the most clearly anomalous melt in the James Ross Island core occurs after the late 1970s, about the time the ozone hole appeared. But the melt data also show that melting has increased nearly monotonically since the 1930s, well before the advent of the ozone hole. As in West Antarctic δ¹⁸O, there was a bit of an increase in melt in the 1830s and the 1940s at James Ross Island, perhaps also ENSO-related, but these little bumps pale in comparison with the amount of melting occurring since the 1950s.

So what does all this mean for the fate of Antarctic Peninsula glaciers and the West Antarctic ice sheet? Both our paper and the Abram et al. paper add substantial new evidence that something rather unusual is occurring in Antarctica. It is not just happenstance that rapid ice sheet, glacier, and ice shelf changes are occurring now, when we have finally begun to observe them closely. Rather, these changes are occurring along with what is happening to the rest of the planet. That said, it appears that we not have yet driven West Antarctic climate (nor West Antarctic glaciers) definitively beyond what might be expected from natural variability alone. In particular, I won’t be surprised if continued decade-to-decade variability in atmospheric circulation results in more, and less, intrusion of circumpolar deep water onto the continental shelf, and to more, and less, rapid thinning of ice shelves in West Antarctica*. On the Peninsula, though, it seems very clear that we have already pushed the system well beyond “normal”, and into conditions reminiscent of the mid-Holocene. I don’t think we’re going to see a return to “normal” conditions any time soon. It’s worth noting that most model projections suggest that the SAM trend may level off for a while as the ozone hole gradually declines, but those same model projections suggest the SAM trend will recover as CO₂ continues to rise. See. e.g. Thompson et al. (2011).

The real take home message here is that the ice loss from the WAIS and from the Antarctic Peninsula that have been observed in the last few decades are indeed likely to be harbingers of things to come. The very rapid rate of change in West Antarctica that we’ve seen over the last few decades is clearly overprinted by substantial decadal variability, so caution is in order in projecting that rate forward in time. The magnitude of the century scale trend will depend quite a bit, in my view, on what happens in the tropics over the next century. The sign of the trend, however, is clear. On the Peninsula, it’s crystal clear.

Note: An excellent summary of these two papers by Tas van Ommen will appear in Nature Geoscience in the May issue.

Guest post by Geert Jan van Oldenborgh, Francisco Doblas-Reyes, Sybren Drijfhout and Ed Hawkins

Climate information for the future is usually presented in the form of scenarios: plausible and consistent descriptions of future climate without probability information. This suffices for many purposes, but for the near term, say up to 2050, scenarios of emissions of greenhouse gases do not diverge much and we could work towards climate forecasts: calibrated probability distributions of the climate in the future.

This would be a logical extension of the weather, seasonal and decadal forecasts in existence or being developed (Palmer, BAMS, 2008). In these fields a fundamental forecast property is reliability: when the forecast probability of rain tomorrow is 60%, it should rain on 60% of all days with such a forecast.

This is routinely checked: before a new model version is introduced a period in the past is re-forecast and it is verified that this indeed holds. In seasonal forecasting a reliable forecast is often constructed on the basis of a multi-model ensemble, as forecast systems tend to be overconfident (they underestimate the actual uncertainties).

As the climate change signal is now emerging from the noise in many regions of the world, the verification of regional past trends in climate models has become possible. The question is whether the recent CMIP5 multi-model ensemble, interpreted as a probability forecast, is reliable.

As there is only one trend estimate per grid point, necessarily the verification has to be done spatially, over all regions of the world. The CMIP3 ensemble was analysed in this way by Räisänen (2007) and Yokohata et al. (2012). In the last few months three papers have appeared that approach this question for the CMIP5 ensemble with different methodologies: Bhend and Whetton (2013), van Oldenborgh et al. (2013) and Knutson et al (J. Climate, to appear).

All these studies reach similar conclusions. For temperature: the ensemble is reliable if one considers the full signal, but this is due to the differing global mean temperature responses (Total Climate Responses, TCR).

When the global mean temperature trend is factored out, the ensemble becomes overconfident: the spatial variability is too low. For annual mean precipitation the ensemble is also found to be overconfident. Precipitation trends in 3-month seasons have so much natural variability compared to the trends that the overconfidence is no longer visible.

These conclusions match with earlier work using the Detection and Attribution framework showing that the continental-averaged temperature trends can be attributed to anthropogenic factors (eg Stott et al, 2003), but zonally-averaged precipitation trends are not reproduced correctly by climate models (Zhang et al, 2007).

The spatial patterns for annual mean temperature and precipitation are shown in figure 1 below. The trends are defined as regressions on the modelled global mean temperature, i.e., we plot B(x,y) in

(1) T(x,y,t) = B(x,y) T_global,mod(t) + η(x,y,t)

This definition excludes the TCR and minimises the noise η(x,y,t) better than a trend that is linear in time.

Figure 1: Panels a and b show the trend in annual mean GISTEMP temperature analysis and the GPCC precipitation analysis over 1950–2011 defined by Eq.(1). Panels c and d show the same for the CMIP5 multi-model mean (historical+RCP4.5). Panels e and f show the percentile of the observed trend in the CMIP5 ensemble of trends. Coloured areas denote where the observed trend is in the tails of the ensemble. Panels g and h collect these percentiles (north of 45ºS) in rank histograms. The top and bottom 5% should only occur 5%, but in fact two to four times more of the map is in these percentiles. The grey area in the rank histograms denotes the 90% confidence inerval.

The conclusion that the ensemble is somewhat overconfident is based on the bottom two panels. These show that over 10%–20% of the maps the observed trends are in the top and bottom 5% of the ensemble. For a reliable ensemble this should be 5%. The deviations are larger than we obtain from the differences between the models (the grey area).

On the maps above the areas where the modelled trends fall in the tails of the ensemble are coloured. In part of these areas the discrepancies are due to random weather fluctuations, but a large fraction has to be ascribed to forecast system biases. (The results do not depend strongly on the observational dataset used, with HadCRUT4.1.1.0, NCDC LOST and CRU TS 3.1 we obtain very similar figures, see the Supplementary Material of van Oldenborgh et al).

These forecast system biases can arise in three ways.

First, the models may underestimate low-frequency natural variability. Knutson et al show that natural variability in the warm pool around the Maritime Continent is indeed underestimated up to time scales of >10 years, contributing to the discrepancy there in Fig.1e. In most other regions the models have the correct or too large variability.

Another cause may be the incorrect specification of local forcings such as aerosols or land use. As an example, visibility observations suggest that aerosol loadings in Europe where higher in winter in the 1950s than assumed in CMIP5. This influences temperature via mist and fog (Vautard et al, 2009) and other mechanisms.

Finally, the model response to the changes in greenhouse gases, aerosols and other forcings may be incorrect. The trend differences in Asia and Canada are mainly in winter and could be due to problems in simulating the stable boundary layers there.

To conclude, climate models can and have been verified against observations in a property that is most important for many users: the regional trends. This verification shows that many large-scale features of climate change are being simulated correctly, but smaller-scale observed trends are in the tails of the ensemble more often than predicted by chance fluctuations. The CMIP5 multi-model ensemble can therefore not be used as a probability forecast for future climate. We have to present the useful climate information in climate model ensembles in other ways until these problems have been resolved.

References

1. T.N. Palmer, F.J. Doblas-Reyes, A. Weisheimer, and M.J. Rodwell, "Toward Seamless Prediction: Calibration of Climate Change Projections Using Seasonal Forecasts", Bulletin of the American Meteorological Society, vol. 89, pp. 459-470, 2008. http://dx.doi.org/10.1175/BAMS-89-4-459
2. J. RÄISÄNEN, "How reliable are climate models?", Tellus A, vol. 59, pp. 2-29, 2007. http://dx.doi.org/10.1111/j.1600-0870.2006.00211.x
3. T. Yokohata, J.D. Annan, M. Collins, C.S. Jackson, M. Tobis, M.J. Webb, and J.C. Hargreaves, "Reliability of multi-model and structurally different single-model ensembles", Climate Dynamics, vol. 39, pp. 599-616, 2012. http://dx.doi.org/10.1007/s00382-011-1203-1
4. J. Bhend, and P. Whetton, "Consistency of simulated and observed regional changes in temperature, sea level pressure and precipitation", Climatic Change, vol. 118, pp. 799-810, 2013. http://dx.doi.org/10.1007/s10584-012-0691-2
5. G.J. van Oldenborgh, F.J. Doblas Reyes, S.S. Drijfhout, and E. Hawkins, "Reliability of regional climate model trends", Environmental Research Letters, vol. 8, pp. 014055, 2013. http://dx.doi.org/10.1088/1748-9326/8/1/014055
6. P.A. Stott, "Attribution of regional-scale temperature changes to anthropogenic and natural causes", Geophysical Research Letters, vol. 30, 2003. http://dx.doi.org/10.1029/2003GL017324
7. X. Zhang, F.W. Zwiers, G.C. Hegerl, F.H. Lambert, N.P. Gillett, S. Solomon, P.A. Stott, and T. Nozawa, "Detection of human influence on twentieth-century precipitation trends", Nature, vol. 448, pp. 461-465, 2007. http://dx.doi.org/10.1038/nature06025
8. R. Vautard, P. Yiou, and G.J. van Oldenborgh, "Decline of fog, mist and haze in Europe over the past 30 years", Nature Geoscience, vol. 2, pp. 115-119, 2009. http://dx.doi.org/10.1038/ngeo414

Some of my friends have made a film, Thin Ice, which tells the story of CO2 and climate from the standpoint of the climate scientists who are out there in the trenches trying to figure out what is going on. I have a small role in the film myself, and I am sure RealClimate readers will recognize many more familiar faces. One of the many things I like about this film is that it puts a human face on climate science. It’s harder to demonize people when you feel you know them, and realize that in the end they’re not that different from you and your neighbors (except maybe they know more about CO2 and climate than some others you might meet).

A description of the project, including trailers and clips can be found here . The film will be available during Earth Week for free streaming. Or even better, you can arrange a free screening for your group (details for obtaining a free Earth Week download for screening are available here ). Read below the fold for more information

Here is what Peter Barrett, the team leader for the film project has to say:

“A group of us have produced another film about climate science, but in this one scientists do the talking.

Some are well known to you, others not so, but all talk with passion , concern and some humour about their work. The film is mainly the work of geologist and photographer Simon Lamb and science documentary producer David Sington, DOX Productions, who worked together on Earth Story (BBC Horizon, 1998). The story line is Simon’s journey as a geologist. He has heard the terrible things the press have been reporting about his climate science colleagues, so he decides to take his camera and find out what’s really happening.

The key messages from this 73 minute film are that scientists can be trusted and that ultimately we have to quit using fossil fuels. We do not try and say how this should be done, but we hope that the film will lead audiences into some deeper thinking on the issue and perhaps even a shift toward solutions. Check out the website www.thiniceclimate.org , where you can see the 3 minute trailer. The website contains another 3 hours of supplementary material in 37 short video clips about various aspects of climate science.

We’d like your help in spreading these messages by hosting a screening in your community. It’s also a chance to talk with them afterwards through a panel discussion/Q&A. We are making the film available as a free download (2GB) for a 2 ½ day period after Earth Day starts in New Zealand – just complete the Screenings Information sheet attached and e-mail to thiniceclimate@vuw.ac.nz so we can post it on the website and send you download instructions. The film will also be available free for streaming to those who are happy just to watch it at home.

We’ve also attached a one-pager on the project and a poster

Feel free to pass this message on. Looking forward to hearing from you soon.

Peter Barrett for the Thin Ice Team

PS While the film is in English with a range of accents we’ll have versions with subtitles in English, Mandarin, Spanish, French and German.”