RealClimate logo

Ice-core dating corroborates tree ring chronologies

Filed under: — group @ 5 August 2015

Guest commentary from Jonny McAneney

You heard it here first

Back in February, we wrote a post suggesting that Greenland ice cores may have been incorrectly dated in prior to AD 1000. This was based on research by Baillie and McAneney (2015) which compared the spacing between frost ring events (physical scarring of living growth rings by prolonged sub-zero temperatures) in the bristlecone pine tree ring chronology, and spacing between prominent acids in a suite of ice cores from both Greenland and Antarctica. The main conclusion was that ice core dates, in particular those ice cores relied upon the Greenland Ice Core Chronology 2005 (GICC05), such as the NEEM S1 core, were too old by approximately seven years during the 6th and 7th centuries AD.

Last month, in an excellent piece of research (Sigl et al., 2015) by a collaboration including Earth scientists, dendrochonologists, and historians, the chronology of the Greenland North Eemian Ice Drilling core (NEEM) has been reassessed and re-dated, confirming that such an offset does indeed exist in the GICC05 timescale below AD 1000. The clinching evidence was provided by linking tree-ring chronologies to ice cores through two extraterrestrial events…

A cosmic connection

In 2012, Miyaki et al. discovered a rapid increase of radiocarbon (14C) in Japanese cedar, precisely dated to AD 774-775. The cause of this increase was possibly due to a very high energy solar proton event (Usoskin et al., 2013), and its effect on radiocarbon has been observed in tree ring chronologies in both hemispheres (Güttler et al., 2015). Another rapid, slightly smaller (~60%), radiocarbon production event has been dated to AD 993-994 (Miyake et al., 2013).

But 14C is not the only cosmogenic isotope produced by such high energy events. Specifically, Beryllium-10 (10Be) is formed from high energy collisions with N and O in the atmosphere, and because of its long lifetime and affinity for soluble aerosols, it precipitates out of the atmosphere quickly and can be measured in ice cores. Therefore, high energy cosmic or solar events should simultaneously create excess 14C and 10Be, and be measurable in tree-rings and ice cores respectively.

By locating the 775 and 993 spikes in 10Be in the ice cores, Sigl et al. (2015) have effectively created chronological tie points to the tree rings with almost zero error. Looking at the 10Be record of the NEEM S1 ice core as well the new TUNU ice core from Greenland, and the West Antarctic Divide ice core (WDC), they found that the peak 10Be deposition was apparently offset from trees by 6-7 years, with the ice core chronologies being too old. These events, as well as tephra markers and historical records of dust veils, were used constrain and evaluate the dating of the ice cores and has led to the creation of a new dating scheme, NEEM NS1 (see figure 1).

Figure 1: New ice core timescale of Greenland ice core NEEM NS1 (top) and Antarctica ice cores (bottom), and the effects of their forcing on a climate reconstruction from a small selection of tree rings (middle) (Sigl et al., 2015).

Historical Implications

This new dating scheme improves understanding of volcanic forcing effects before the 1st millennium. For example, 15 of the 16 coldest summers between 500 BC and AD 1000 follow immediately after volcanic events, four of which are found shortly after the largest volcanic events in the record. It also confirms that the so-called “AD 536 event” was a two-stage event, with two large eruptions, a Northern hemisphere eruption in AD 536, and tropical eruption in AD 540, causing over a decade of poor climate and hardship on human civilisation. As we’ve mentioned previously, these eruptions may even have paved the way for the Justinian plague to take hold after AD 540, which is estimated to have killed about one third of the European population, and may also have impacted the Maya civilisation.

The new chronology also shows that the GICC05 date of 1104 for a volcanic horizon attributed to the historical eruption of the Icelandic volcano Hekla is erroneous. Since this horizon is now dated to AD 1108/9 it cannot be associated with Hekla, explaining why attempts to identify Hekla tephra in ice cores in this layer were unproductive (Coulter et al., 2012). It also confirms that the dating of the historical eruption of Eldgja to 934 is also incorrect, its actual eruption date being AD 939. And of course, the acid and tephra signal which was thought to have originated with the AD 79 Vesuvius eruption (Barbante et al., 2013), cannot possibly be due to Vesuvius, this layer now being dated to AD 88.

The one thing not mentioned though is how the error in GICC05 originated. The ice core dates for Eldgja and Hekla of 934 and 1104 have been in print since 1980 (Hammer et al., 1980), and they have been replicated in each Greenland ice core up to now, culminating in the GICC05 timescale. There are similar offsets with Antarctica ice cores that have been dated independently of the GICC05 timescale. An understanding of the origin of this error could help to evaluate the robustness of all ice core dating.

Future hopes for the past

This research (only!!!) covers the last 2500 years; Half the period of human history (~5000 years). Synchronising ice cores with tree dates, as well as with historical dates, is even more problematic the further in the past one goes. If the ice core chronologies can be extended further into the past with accuracy equal to that of the NEEM NS1 core, the effects of volcanic forcing and climate more generally on past civilisations can be improved. For example, we know that Hekla had two massive eruptions in the past, probably in the 12th and 24th century BC, but we do not know exactly when, or what effect they may have had on climate and society. We do not yet know the cause of the so called 4.2 kya event (from 2200-1900 BC), which was a period of altered climate which may have led to the collapse of the Egyptian Old Kingdom, and/or the rise of the Akkadian Empire. Perhaps sufficiently accurately dated ice cores extending across these event could provide the answer?

Finally, a key controversy in archaeological research is the accurate dating of the Thera eruption, thought to have occurred in the 17th or 16th century BC. Positive identification and accurate dating of tephra in ice cores from this large Mediterranean eruption would provide a critical benchmark for early Middle Eastern history.

Trees can guide us in reconstructing the past. As my co-author Mike Baillie quipped in his book A Slice Through Time: Dendrochronology and Precision Dating,   “The trees don’t lie – and they were there”. Thanks to the research of Sigl et al. (2015), this statement is looking all the more true.


  1. M.G.L. Baillie, and J. McAneney, "Tree ring effects and ice core acidities clarify the volcanic record of the first millennium", Climate of the Past, vol. 11, pp. 105-114, 2015.
  2. B.M. Vinther, H.B. Clausen, S.J. Johnsen, S.O. Rasmussen, K.K. Andersen, S.L. Buchardt, D. Dahl-Jensen, I.K. Seierstad, M. Siggaard-Andersen, J.P. Steffensen, A. Svensson, J. Olsen, and J. Heinemeier, "A synchronized dating of three Greenland ice cores throughout the Holocene", J. Geophys. Res., vol. 111, 2006.
  3. M. Sigl, J.R. McConnell, L. Layman, O. Maselli, K. McGwire, D. Pasteris, D. Dahl-Jensen, J.P. Steffensen, B. Vinther, R. Edwards, R. Mulvaney, and S. Kipfstuhl, "A new bipolar ice core record of volcanism from WAIS Divide and NEEM and implications for climate forcing of the last 2000 years", Journal of Geophysical Research: Atmospheres, vol. 118, pp. 1151-1169, 2013.
  4. M. Sigl, M. Winstrup, J.R. McConnell, K.C. Welten, G. Plunkett, F. Ludlow, U. Büntgen, M. Caffee, N. Chellman, D. Dahl-Jensen, H. Fischer, S. Kipfstuhl, C. Kostick, O.J. Maselli, F. Mekhaldi, R. Mulvaney, R. Muscheler, D.R. Pasteris, J.R. Pilcher, M. Salzer, S. Schüpbach, J.P. Steffensen, B.M. Vinther, and T.E. Woodruff, "Timing and climate forcing of volcanic eruptions for the past 2,500 years", Nature, vol. 523, pp. 543-549, 2015.
  5. F. Miyake, K. Nagaya, K. Masuda, and T. Nakamura, "A signature of cosmic-ray increase in ad 774–775 from tree rings in Japan", Nature, 2012.
  6. I.G. Usoskin, B. Kromer, F. Ludlow, J. Beer, M. Friedrich, G.A. Kovaltsov, S.K. Solanki, and L. Wacker, "The AD775 cosmic event revisited: the Sun is to blame", Astronomy & Astrophysics, vol. 552, pp. L3, 2013.
  7. D. Güttler, F. Adolphi, J. Beer, N. Bleicher, G. Boswijk, M. Christl, A. Hogg, J. Palmer, C. Vockenhuber, L. Wacker, and J. Wunder, "Rapid increase in cosmogenic 14C in AD 775 measured in New Zealand kauri trees indicates short-lived increase in 14C production spanning both hemispheres", Earth and Planetary Science Letters, vol. 411, pp. 290-297, 2015.
  8. F. Miyake, K. Masuda, and T. Nakamura, "Another rapid event in the carbon-14 content of tree rings", Nature Communications, vol. 4, pp. 1748, 2013.
  9. S.E. Coulter, J.R. Pilcher, G. Plunkett, M. Baillie, V.A. Hall, J.P. Steffensen, B.M. Vinther, H.B. Clausen, and S.J. Johnsen, "Holocene tephras highlight complexity of volcanic signals in Greenland ice cores", Journal of Geophysical Research: Atmospheres, vol. 117, pp. n/a-n/a, 2012.
  10. C. Barbante, N.M. Kehrwald, P. Marianelli, B.M. Vinther, J.P. Steffensen, G. Cozzi, C.U. Hammer, H.B. Clausen, and M. Siggaard-Andersen, "Greenland ice core evidence of the 79 AD Vesuvius eruption", Climate of the Past, vol. 9, pp. 1221-1232, 2013.
  11. C.U. Hammer, H.B. Clausen, and W. Dansgaard, "Greenland ice sheet evidence of post-glacial volcanism and its climatic impact", Nature, vol. 288, pp. 230-235, 1980.

23 Responses to “Ice-core dating corroborates tree ring chronologies”

  1. 1
    Chris Ho-Stuart says:

    Does this have any implications for ice core timing beyond the normal range of dendrochronology? Specifically, a common talking point has been the time gap between onset of temperature increase coming out of the last glacial maximum and the start of CO2 rise. It’s been generally accepted for some time that the was an 800 year lag; with the temperature rise starting first and CO2 rise following after.

    In popular debate over AGW this has been used by deniers as proof that CO2 doesn’t cause temperature rise; while conventional science has that CO2 as a feedback can still be responsible for most of the temperature increase of the several thousand years involved. For instance, at real climate Dec 3, 2004: “What does the lag of CO2 behind temperature in ice cores tell us about global warming?”

    Recent research has since called into question the existence of this lag at all: “Synchronous Change of Atmospheric CO2 and Antarctic Temperature During the Last Deglacial Warming” by F. Parrenin et al (in Science 1 March 2013: Vol. 339 no. 6123 pp. 1060-1063 DOI: 10.1126/science.1226368). Parrenin et al are also proposing a revision to time scales of Antarctic ice cores.

    I’m not sure of the state of play in all this at present; and I’m wondering if your work is at all relevant to the issue of time difference between onset of temperature rise and CO2 rise at the end of the most recent glacial.

  2. 2
    Chris Baisan says:

    Great news! Congratulations on the vindication!

  3. 3
  4. 4
    Arthur Smith says:

    Great and clear article. But how do we know the tree rings are right and ice cores are wrong?

  5. 5
    Chris Baisan says:

    Arthur S – we know the tree-ring chronologies are correct because, unlike the ice cores, there are literally hundreds of independently dated times series in both hemispheres that can be cross checked with each other either directly when in the same climatic region, or indirectly as in the case of the radiocarbon spikes discussed here. The first paper describing the spike was based on material from Japan and gave a calendar date. This calendar date has been verified using tree-ring material from other continents. There is a global network of tree ring sites and they all agree – only the ice cores were out of sync and the dating of them has always included error bars – unlike the tree-ring chronologies.

    Chris H- the next logical step is to run the Be analysis on the ice cores back and verify the calendar dates with single year radiocarbon analysis of dated tree rings. This will cover the last 12k years. Further back there are floating sequences of one to several thousand years of crossdated tree rings that cover parts of the last 20k years. That will help with the ice dating, however if the CO2 signal and temperature signal have the same source (same ice core) then there are other issues at work.

  6. 6
    John Pollack says:

    @4 – The tree rings are right, because tree ring series from both hemispheres show a carbon 14 spike at 774-5 CE due to exceptional solar activity. If the tree rings were wrong, one or several of these series would be displaced by several years.

  7. 7
    Greg Simpson says:

    Recent research has since called into question the existence of this lag at all…
    That seems a little odd. Assuming the underlying cause of glaciations is the Milankovitch cycles I don’t see how there could not be a lag. I have no clue about how large it should be, though.

  8. 8

    @Chris Ho-Stuart:
    The fact that the Greenland and Antarctic ice cores (or rather GICC05 and AICC2012) are synchronous for the Holocene is largely due to the temporal flexibility used in the comparison. Veres et al. (2013, Climate of the Past, 9, p. 1739) write: “we used a set of absolute markers from the GICC05 depth-age relationship with artificially small uncertainties to force the model not to deviate from the chronology”. Furthermore, the analysis uses tie-points of 50-year uncertainties for this period, which is probably larger than the actual discrepancy in dates between IntCal and GICC05. When looking beyond the Holocene, the Antarctic ice core dates are younger than the GICC05 equivalents for the same events. By GI-22 and GI-23, it is on the order of millennia (see Veres et al.).

    Some of the further analysis suggested above can be found in the articles I linked to previously.

  9. 9
    t marvell says:

    Re Post #1 – Does this mean that there is now annual data, that is generally considered accurate, for both CO and world temperature going back well over 1000 years? Is it readily available?

  10. 10
    Jonny McAneney says:

    Arthur Smith – Chris Baisan explains scientifically why it is the tree rings that are correctly dated and not the ice cores. But to further his statement, we also know that it is the tree rings that are correct and not teh ice cores because of historical documentation.

    To reiterate a point made in a previous post see (figure 3 here, we have historical records of recorded phenomena consistant with volcanic dust veil events in 44 BC, AD 536 and AD 627. These dates match the formation of frost damaged rings in bristlecone pine in 43 BC, AD 536 and AD 626/7. However when we look at the GICC05, they have acid at 50 +/- 1 BC, AD 529 +/- 2 and AD 619 +/- 2. Of course in fairness they do have acid at around 534 AD, which Larsen et al 2008 argued was the cause of the AD 536 dust veil event, which we now know to be the AD 540 eruption.

    The nature paper of Sigl et al. 2015 uses the radiation events of 774/5 and 993/4 to tie trees to ice cores creating two synchronisation points. One could again question if the trees are telling the truth with respect to their dates. By comparing the onset of the radiocarbon signal between northern and southern hemispheric trees the timing of the 774-775 event has been refined to be March 775 +/- 6 months (Guttler et al., 2015). This timing agrees quite well with dated Chinese records (see Zhou et al. 2013 and references within But here is the pertinant section

    “In the evening on the Chinese lunar calendar day of 11
    Dec. 774, i.e., 17 Jan. AD775, in the east and above Moon,
    there were more than ten bands of white lights like the
    spread silk, penetrating and covering eight grand constellations
    named in Chinese, corresponding to the sky composed
    of Taurus-Auriga, m Gemini, q Cancer, l Orion,
    V Orion, e Taurus, d Hydra and a Leo, and the lights
    were ceased gradually after middle night, as recorded in
    the Old Tang Book – a Chinese Chronicle [7]. The auroras
    described are actually located in a wide region from east
    to west and from north to south in Earth’s northern hemisphere.
    The records indicated clearly that time to see auroras
    is most probably from 5:00-6:00 PM to 1:00-2:00 AM,
    therefore, the event lasted about 8 hours. In Tang dynasty,
    the natural abnormal events were observed and recorded by
    qualified officers, so the records are highly trustable and,
    the work was conducted in the Tang capital Xian, which
    is located in a geomagnetic latitude at lower twenties. The
    low geomagnetic latitude indicates that the solar particles
    which caused very strong auroras are very intense and with
    higher energies.”

    This might not seem that remarkable until one considers that this observation was recorded in the Tang capital Xian which is around the 34th parallel, implying that this was an extremely energetic auroral event.

  11. 11
    Jonny McAneney says:

    Chris Ho-Stuart – I am not a climate scientist so I cannot answer definatively your question, but I will give it a shot within my own limited understanding. If I make any errors, hopefully others will point it out.

    My understanding is that much of the temperature reconstruction of the past is obtained from measurement of d18O within ice core strata. Similalry past CO2 concentration is also reconstructed from ice core layers. Thus, one would assume that if the measurements were taken from the same core, then absolute dating does not matter too much with respect to any correlation between CO2 and temperature, since one is dealing only with that core’s internal chronology. Issues would probably arise if you are comparing proxy data between different ice cores, which would depend upon the degree of established synchroniety between cores.

  12. 12
    Jonny McAneney says:

    Chris Baisan – Thank you for your responses to the other posters. I just wanted to enphasis a point you make regarding syncronising tree-rings with ice cores using 14C and 10Be. As of 2102/2013 Miyake et al. had not identified any further rapid increases in 14C, other than AD 774/5 and AD 993, within the previous three millennia, but I am sure that this wont stop them, or others, looking for further excess radiocarbon signals within tree ring records.

    If I remember correctly Miyake et al. were guided to the 775 and 993 events through the radiocarbon calibration curve, which prompted them to do high precision measurements (i.e. at annular resolution) 14C measurements around these dates. Similaly for 10Be measures in ice cores, where annular resolution measurements were taken because of the radiocarbon measurements. It might be a tall order to obtain annular resolution of radiocarbon from trees and 10Be from ice cores for the twelve millennia of tree ring data. Radiocarbon and 10Be resolution is currently at five to ten year resolution (perhaps even coarser for 10Be, I cant recall at the moment).

    There is no doubt that complete annularly resolved 14C and 10Be chronologies would be highly desirable, but obtaining them might be quite a costly endeavour.

  13. 13
    Chris Baisan says:

    Yes, Jonny, a strategy is required. A matter of determining the most efficient way to target portions of the radiocarbon calibration curve/10Be numbers. There are many reasons to want single year numbers across the whole span of the RC calibration curve but obtaining financing is a stumbling block for sure.

  14. 14
    Richard Caldwell says:

    I just want to say that I’m amazed that ice cores can resolve to even a decade. Snow and stuff mixes together as ice forms. I tip my hat to scientists for even trying to solve this issue. I’d probably just declare tree rings right and move on. Perhaps we’re smarter than we think. Perhaps we’ve got a chance. Folks are generally self-absorbed and self-interested, but folks are also generally good.

  15. 15

    Fascinating article. This added empirical evidence of temperatures in the past will help to silence the fossil fuel industry funded critics of climate change science.

    As a side issue, does this mean Carbon 14 dating will require some equation adjustments due to the identification of high proton extra terrestrial events?

  16. 16
    Chris Baisan says:



    As has been mentioned above the calibration curve is derived mostly from 10 year increments (and some 5 year) which introduces a smoothing error but mitigates noise. At the time calibration work was begun the mass required for each increment was considerable and testing single years wasn’t practical in most cases. Now the mass required is no longer a barrier because of technical advances but the cost is a barrier. The effect of the ‘wiggles’ in the curve is that the potential error or uncertainty of a radiocarbon date varies through time – not from counting or measurement error which is also a factor – so in some periods the resolution of the method is poor, in others very good. The spikes don’t have a simple effect because near the spike (year of or just following)the dating error is potentially very large while either side of the effect the error is reduced – if the spike is included/accounted for in calibration curve. This applies to ‘typical’ organic samples, not tree-ring samples with annual layers. Overall effect of identifying and accounting for spikes should be a reduction of dating error generally with some specific exceptions. An ‘unknown’ sample with annual layers that include the spike could theoretically be dated with uncertainty almost eliminated if one does dates for multiple layers. I am not sure how the calibration people will handle this but there is active interest in the subject and it will be addressed in some fashion.

  17. 17

    RC: folks are also generally good.

    BPL: Explain how generally good folks produced Auschwitz, the GULAG, and a world where 1/9 of the people are starving.

  18. 18
    Hank Roberts says:

    Are you folks talking with the researchers working with organic lipids in sediments? I know this is relatively new compared to C14 but I’ve seen quite a few mentions. Just a random ‘oogle turns up e.g.

    Biogeosciences, 7, 3473–3489, 2010
    © Author(s) 2010. CC Attribution 3.0 License.

    Lipid biomarkers in Holocene and glacial sediments from
    ancient Lake Ohrid (Macedonia, Albania)
    Published in Biogeosciences Discuss.: 16 June 2010
    Revised: 7 October 2010 – Accepted: 12 October 2010 – Published: 8 November

    Variable proportions of terrestrial n-alkanoic acids and n-alkanols as well as compositional changes of ω-hydroxy acids document differences in soil organic matter supply between the sites and during different climate stages (glacial, Holocene, 8.2 ka cooling event). Changes in the vegetation cover are suggested by changes in the dominant chain length of terrestrial n-alkanols. Effective microbial degradation of labile organic matter and in situ contribution of organic matter derived from the microbes themselves are both evident in the sediments. We found evidence for anoxic conditions within the photic zone by detecting epicholestanol and tetrahymanol from sulphur-oxidising phototrophic bacteria and bacterivorous ciliates and for the influence of a settled human community from the occurrence of coprostanol, a biomarker for human and animal faeces (pigs, sheep, goats), in an early Holocene sample. This study illustrates the potential of lipid biomarkers for future environmental reconstructions using one of Europe’s oldest continental climate archives, Lake Ohrid.

    I recall papers on marine sediments as well.

  19. 19
    mitch says:

    For A.G. Gelbert–
    C-14 dates have been adjusted for the non-constant production rate for a while. Since the 1980’s at least, researchers have compared ‘radiocarbon age’ assuming constant production rate to calendar age through tree ring counts. There are on-line programs to correct your radiocarbon data (

    Anything that improves the scale and provides cross-calibration helps to narrow down errors and improves our view of past events.

  20. 20

    “BPL: Explain how generally good folks produced Auschwitz, the GULAG, and a world where 1/9 of the people are starving.”

    – See more at:

    Geez, Barton, don’t throw us any softballs…

  21. 21
  22. 22

    #21–Great, but to do that I’d have to actually read the damn thing. Life is short.

  23. 23
    John Mashey says:

    Nice work, some of us appreciate the sometimes-thankless tasks of resolving data discrepancies, especially in paleo problems where one cannot just return experiments in the lab.

Switch to our mobile site