Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 1 of 13
Possible errors in the Eastern Alpine Conifer Chronology
Petra Ossowski Larsson* and Lars-Åke Larsson, Sweden
* Corresponding author: email@example.com
The Eastern Alpine Conifer Chronology is clearly synchronized with the European
oak chronologies over the recent 2500 years, thus confirming the long established
dendrochronological bridge over the "Roman gap" which we dispute. We claim that
the European timber complex archaeologically anchored in Roman time is
conventionally dated too old by 218 years. But as the raw measurement data of the
Alpine chronology is unpublished and unavailable we can not check whether our
hypothesis is wrong, or the chronology is in error. However, some "outliers" in data
derived from the chronology seem to tilt the scales in our favor.
The principle of dendrochronology was first formulated by Andrew E. Douglass in
1919, and later introduced in Europe by the botanist Bruno Huber. But it was Ernst
Hollstein who systematically established an absolutely dated "standard oak tree-ring
chronology" for western Germany as an aid for age determination of archaeological
artefacts. Oak was the preferred construction wood for both the Romans and later
builders. By 1965, this tree-ring chronology reached continuously back to 739. About
that time Hollstein set himself the target to reach the already well established but still
floating oak tree-ring sequences archaeologically anchored in Roman time by
bridging Early Medieval time. However, this task turned out to be more difficult than
expected because of the lack of suitable timber. Especially difficult was the traversing
of two gaps, one at 750 and a wide one between 300 and 400. The work was
regarded as reasonably completed in 1975 and the first long oak chronology was
published in its entirety in 1980 (ref.1).
Only few years later in 1984, a continuous oak tree-ring chronology for western
Europe which spanned more than 7000 years was announced as completed in an
article (ref.2). It was a joint venture by the dendro-labs in Belfast, Köln and Stuttgart
Hohenheim, involving oak chronologies from the north of Ireland, northern Germany
and southern Germany. This joint venture was necessary because neither the Irish
nor the German chronologies were at that time stand-alone. The weakest link in the
Irish chronology was then thought to be at 250 BC, and in the German chronology at
550 BC. At that point had actually been an error and 71 extra years were inserted in
the German chronology. Furthermore, in order to connect the BC chronology to the
absolute AD chronology, additional material had been necessary. In the Belfast case,
the English chronologies from Carlisle and Southwark were used as a bridge. In
Germany, the link between late BC/Roman time and Early Medieval time was
reinforced by the West German chronology of Ernst Hollstein.
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 2 of 13
The new long oak chronologies were immediately used to build the atmospheric part
of the radiocarbon calibration curve (IntCal) still in use (ref.3). Since then the long
European oak chronologies have been revised and amended, but their year count for
the past 7000 years has never been changed and all are perfectly synchronized with
each other. A lot of new regional chronologies have been developed as well. This
means that Hollstein's original bridges to Early Medieval time and then to Roman
time are still regarded as valid as they were in 1975. However, as there was a lot of
trouble with these bridges (both had to be corrected after the initial announcement)
and also with some of the datings which were obtained from the finished chronology,
people including Hollstein himself suspected errors.
With effective dendrochronological tools right at our hands, we started in 2006 a
project to validate Hollstein's crucial bridges, i.e. to demonstrate with
dendrochronology the continuity of the existing long European tree-ring chronologies
back into Roman time (ref.4).
As a first result, we could verify that the gap at 750 was correctly bridged by
Hollstein. We were however not able to validate his bridge over the "Roman gap" as
there is a weak period of about two hundred years in virtually all available European
oak chronologies between Roman time and Early Medieval time. Moreover, because
of difficulties to reliably bridge this gap it had apparently been necessary to "calibrate"
the Roman complex of the Middle European oak chronology with historical
considerations in order to get a useful (i.e. historically correct) dating tool (ref.5).
This result was not expected by us and definitely not suitable for the validation of the
long European oak chronologies. We gave the case a second try after we had
confirmed that two supra-long Scandinavian pine chronologies are continuous and
therefore represent a true dendrochronological projection of the real time line. We
then could demonstrate a significant correlation between these pine chronologies
and a long continuous but floating north-west European oak chronology anchored
archaeologically in Roman time. This exercise showed that the whole Roman oak
complex most probably is conventionally dated too old by 218 years (ref.4).
The here presented investigation deals with the properties of another supra-long
European conifer chronology, the more than 9000 years long Eastern Alpine Conifer
Chronology, in order to shed more light on this peculiar result.
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 3 of 13
Properties of the Eastern Alpine Conifer Chronology (EACC)
This 9111 years long compound chronology (built from Swiss stone pine (pinus
cembra), larch (larix decidua) and Norway spruce (picea abies)) from the western
section of the central Eastern Alps on both sides of the main Alpine ridge was
announced by Kurt Nicolussi et al. in 2009 (ref.6). Its expressed purpose was to
provide at least annual resolution for investigations of Holocene climate variability in
a high mountain region which had already proved particularly well suited for
paleoclimatic studies. But of course it also provides a dating reference for
Although several early efforts had been made to establish a continuous Holocene
tree-ring chronology in the Alps, the absolute part of the EACC (anchored in living
trees) reached back only to Early Medieval time until about 2000. However, already
at the end of 2002 the absolute chronology had been extended to more than 7000
years length (ref.7). In that article the direct synchronization with Becker's south
German oak chronology (ref.8, the precursor of the Hohenheimer Jahrringkalender)
is described, meaning that the EACC is also synchronous with the Hollstein
chronology and all other long European oak chronologies over the recent 2500 years.
The complete mean tree-ring width curve and the raw measurement data of the
EACC are unpublished and unavailable. However, in 2011 Büntgen et al. (ref.9)
published a summer temperature reconstruction for the recent 2500 years based on
the EACC among other chronologies. The supplementary material of this article also
includes the mean value tree-ring width curves for the examined time range.
Observations and discussion
The 2500 years long published mean value tree-ring width curve of the EACC
allowed us to check its synchronization with the European oak chronologies. This
check showed that the EACC clearly confirms the conventional dendrochronological
link between Early Medieval time and Roman time as originally proposed by Ernst
Hollstein. This result means either that we are wrong when we propose a different
position of the Roman complex on the real time line, or that the EACC is in error. A
dendrochronological re-analysis of the raw measurement data could solve the case,
but this data is unavailable.
Instead we have to look for indications in the publications about the EACC. The
"Material and methods" section of ref.6 provides some important clues:
These ‘mean tree’ sample series were used in comparisons with other tree-
ring series and reference chronologies. We used visual and statistical
comparisons between single series and already established chronologies to
find the exact crossdating position for each single tree-ring series. Visual
and statistical comparisons were performed after high-pass filtering of the
original measurement series. We applied a 30 yr spline (Cook and Peters,
1981) to remove age-related growth trends from the series. Additionally,
radiocarbon dating was applied extensively mainly at the beginning of
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 4 of 13
chronology building to establish in a first step a framework of approximately
dated tree-ring series and chronologies [ref.7]. In the latter phases of the
project radiocarbon dating was mainly used on samples without satisfactory
And there are more clues in ref.7:
Der schnelle Aufbau der ostalpinen Zirben-Chronologie wurde durch die
14CDatierung einer Reihe von Holzproben wesentlich unterstützt. [ ... ] Die
14C-Alter wurden jeweils auf der Basis der Kalibrierkurve INTCAL98 (Stuiver
et al., 1998) in Kalenderjahre umgerechnet. Ein Vergleich der kalibrierten
14C-Daten und der dendrochronologisch bestimmten Probenalter zeigt eine,
abgesehen von wenigen Ausreißern, klare Übereinstimmung und bestätigt
die Genauigkeit der Jahrringchronologie.
(The rapid development of the stone pine chronology for the eastern Alps
was considerably supported by the radiocarbon dating of a series of wood
samples. [ ... ] The 14C ages were converted into calendar years on the
basis of the calibration curve INTCAL98 (Stuiver et al., 1998). A comparison
of the calibrated 14C data and the dendrochronologically determined sample
age shows a clear agreement, apart from a few outliers, and confirms the
accuracy of the tree-ring chronology.)
Sorry to say, but this description does not make a convincing impression about an
independent development of the chronology . Here we learn that "comparisons with
reference chronologies" and "extensive application of radiocarbon dating" even on
"samples without satisfactory crossdating results" were included in the standard
repertoire when developing the EACC. To this comes that the authors were
apparently aware of the match towards the Becker master. Finally the word "outlier"
So let's see if we can get hold of some of these outliers.
That a comparison of the calibrated 14C data and the dendrochronologically
determined sample age shows a clear agreement does not confirm the accuracy of
the tree-ring chronology. It only confirms that the EACC is synchronized with the tree-
ring chronologies used to build the radiocarbon calibration curve, which we already
know from dendrochronology alone. In other words, the statement in ref.7 is a classic
case of circular reasoning.
However, even if we suspect that a dendrochronological misdating of the Roman oak
complex could have infected also the radiocarbon calibration curve, we know that the
recent two millennia of the calibration curve are based on correct dendrochronology.
This is because the curve between year 1950 and 100 is solely derived from Irish oak
(which is absolute and continuous back to year 25 according to our own
investigations, ref.4), and North-American sequoia and Douglas fir (which is derived
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 5 of 13
from or crossdated with long series from very old trees completely independent of
any historical considerations, ref.10). So, if there would be any dendrochronological
problems in the EACC before year 500, they might reveal themselves as radiocarbon
outliers in the year range 500 to 100.
To visualize this, we calculated the dendrochronologically derived absolute mean
ages of all EACC samples between year 700 and -400 which were radiocarbon
measured and listed in ref.6 (see Table 1). We then plotted these values and their
radiocarbon dates together with their uncertainties against the IntCal13 calibration
curve for this time range (see Figure 1).
Absolute age of
mean age of
SSM-57 637–743 AD 37–56 684 AD
G-29 659–883 AD 1–12 666 AD
GP-107 616–757 AD 38–50 660 AD
G-25 625–916 AD 19–43 656 AD
G-17 546–888 AD 73–92 629 AD
GP-21 244–330 AD 47–65 300 AD
G-1 224–628 AD 1–31 240 AD
G-2 159–667 AD 1–10 165 AD
G-35 130–335 AD 22–36 159 AD
GP-20 53–328 AD 1–17 62 AD
GP-60100 2 BC–164 AD 33–57 43 AD
SSM-20 230–85 BC 138–143 90 BC
LZS-10 142–47 BC 35–47 101 BC
SSM-48 372–168 BC 62–67 308 BC
Table 1: The conventional 14C-dates of samples used for the establishment of the EACC. All data
according to ref.6, table 2, except column "Absolute mean age of 14C sample" which was calculated
by us. All samples Pinus cembra.
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 6 of 13
Figure 1: IntCal13 atmospheric calibration curve (Reimer et al. 2013) for the year range -400 to 800,
visualized with OxCal v4.2.4 (Bronk Ramsey 2013). Compared with the 14C-dates / absolute mean
ages of 14C samples used for the establishment of the EACC, see Table 1.
There is at least one obvious outlier (GP-21) which could hint at problems with the
dendrochronological crossdating within the EACC. But as said above, the raw data
which would facilitate a dendrochronological reassessment is unavailable.
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 7 of 13
Summer temperature reconstruction
A first summer temperature reconstruction for the recent 2500 years based on the
EACC was published in 2011 by Büntgen et al. (ref.9). This temperature
reconstruction has henceforth been used for climate research in several following
publications (e.g. refs.11 to 16).
A graphic from one of the very recent publications (ref.15) caught our interest. Here
the summer temperature constructions for the Russian Altai (a) and the EACC (b) are
directly compared to each other. The curves show a clearly opposed trend between
year 500 and 300.
Figure 2: Upper part from ref.15, figure 2: Eurasian summer temperature variability. a,b,
Reconstructed temperatures from the Russian Altai (a) and the European Alps (b), with blue and
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 8 of 13
red squares indicating (51/47) positive and (53/57) negative annual extremes (>2 standard deviations),
respectively. Grey background shadings denote reconstruction uncertainty after 80-year low-pass
filtering. Black dashed lines refer to the long-term reconstruction mean of the Common Era.
Lower part from ref.17, figure 6: Northern Scandinavian summer temperatures derived from six tree-
ring based reconstructions from northern Sweden and Finland (c). All records were first smoothed
using a 30-year moving average and then calibrated against regional JJA temperatures over the
1889–1979 period using OLS regression.
We added to the graphic the curves of various Scandinavian summer temperature
reconstructions from ref.17 (c) which in the range 500 to 300 rather seem to follow
the trend of the Altai curve than the trend of the curve derived form the EACC.
We have not checked the dendrochronology behind the Altai reconstruction, but we
know from our own investigations (ref.4 and forthcoming results) that the
dendrochronology behind all four Scandinavian reconstructions reaching back to year
1 is sound.
An article of the PAGES 2k Consortium from 2013 (ref.12) finally hints that something
actually might be wrong with the summer temperature reconstruction derived from
the EACC. In this article the temperature variability during the past two millennia is
discussed per continent.
Figure 3: From ref.12, figure 2: Continental-scale temperature reconstructions. 30-year-mean
temperatures for the seven PAGES 2k Network regions, standardized to have the same mean and
standard deviation over the period of overlap among records (ad 1190–1970). North America includes
a shorter tree-ring-based and a longer pollen-based reconstruction. Dashed outlines enclose intervals
of pronounced volcanic and solar negative forcing since ad 850. Data are listed in Supplementary
Database S2. Completed by us with the section older than year 500 of the European reconstruction
cut out and slid 210 years towards younger time corresponding to our proposed dendro-error for the
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 9 of 13
The EACC-based reconstruction is included in "Europe", in the range year 500 and
older accompanied only by a reconstruction based on tree-rings from northern
Finland. In the same time range, the "Artic" is represented by four reconstructions
based on tree-rings from central Russia and northern Scandinavia. The opposed
trend between year 500 to 300 (as noted in figure 2 above), warming in the north and
cooling in the Alps, is also apparent in figure 3 when compared between Arctic and
Europe. When slid 210 years towards younger times, the unique coldest (blue)
European thirty-years period around year 350 will become aligned with the coldest
period around 550 in the same region. Is maybe the blue outlier at 350 only a
mirrored image of the event at 550, due to a dendrochronological error of 218 years
in the EACC?
The Eastern Alpine Conifer Chronology appears to be synchronized with the
European oak chronologies over the recent 2500 years, i.e. the chronology supports
the consensus bridge from Early Medieval time to Roman time as originally proposed
in 1975 by Ernst Hollstein. We have previously found the Hollstein bridge being
dendrochronologically insufficient and propose that the Roman oak complex is
dendro-dated too old by 218 years. As the raw measurement data for the EACC is
unavailable, we can not check if our hypothesis is wrong or if the EACC is in error.
However, we have found irregularities in data derived from the EACC which support
Why then do radiocarbon investigations fail to reveal such a large error in the
European dendro-masters? Most probably this is due to a quite low temporal
resolution of the radiocarbon method which has to rely on a calibration curve. Minze
Stuiver concluded already in 1982 (ref.10) that dendro errors might be responsible for
larger offsets between data sets :
The interlaboratory comparison with La Jolla and Heidelberg yields offsets
between data sets. These offsets (up to 58 radiocarbon years) are most likely
due to laboratory bias. A "real" offset, of perhaps 23 years, appears possible
for the California Sequoia and German Oak radiocarbon ages. This offset
may be due to differences in wood 14C content, but may also be due to
errors in the dendro-age determinations.
This suspicion made that Stuiver avoided risks and therefore rejected data derived
from both bristlecone pine and German oak for the recent two millennia of the
calibration curve. The part of the calibration curve which we suspect is based on
corrupt dendro-data (largely the first millennium BC), contains two large "plateaux"
where a wide range of dendro-years yield a small range of radiocarbon ages. This
complicates dating, but it also effectively conceals any errors in the calibration curve.
It is only if the absolute sample age is surefire known and the measured radiocarbon
value hits one of the steep parts of the calibration curve that any outliers would
become embarrassing apparent. We know of only one such case.
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 10 of 13
Figure 4: IntCal13 atmospheric calibration curve (Reimer et al. 2013) for the first millennium BC
showing two plateaux. One minor plateau with about 100 14C-years per 200 dendro-years, and the
Hallstatt plateau with about 170 14C-years per 400 dendro-years. Below and above the Hallstatt
plateau are two very steep sections of the curve, corresponding to 100 and 160 14C-years per about
20 dendro-years. The measured 14C values of the bone collagen samples from Nineveh (ref.18) with
error margins appear almost entirely in the steep section above the Hallstatt plateau, thus giving a
calibrated age of 800 BC instead of expected ca. 612 BC. Note that the figure has calibrated time
scales for both years BC (top) and years BP (bottom).
The ancient city of Nineveh (near Mosul in Iraq) was sacked by Babylonians and
Medes in -611, an event which is regarded as absolutely dated on the astronomical
timeline via Babylonian clay tablets. Some of the city's last defenders died in the
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 11 of 13
gateway they tried to hold and were instantly buried by the collapsing roof of the
gateway. Their remains were found in situ in 1990.
Radiocarbon analysis of a couple of bone collagen samples from these remains gave
almost all calibrated dates which were about 200 years too old (ref.18). The article by
R.E. Taylor et al. includes a thorough discussion where a number of factors which
could contribute to this large offset are ruled out because the sample preparation was
carefully conducted. However, a heavy fish diet remains as a possible explanation for
As we see it, an alternative to extensive feasting on imported boned, dried and salted
codfish (Bacalhau), which could explain the lack of fish bones in the kitchen middens,
would be an error in the radiocarbon calibration curve.
In the light of our combined results we call for caution when using the EACC for
climate reconstructions, especially before the year 500. To accentuate it once again:
no homogeneous European tree-ring chronology has ever been published which can
demonstrate a significant dendrochronological bridge between Early Medieval timber
and timber archaeologically anchored in Roman time. "Published" means in our
opinion that the raw measurement data is readily available for reassessment to
everybody. Until this problem is properly attended to, we moreover advise scientists
to avoid synchronization of measured natural markers with Roman (and Greek)
historical events, as such a synchronization might turn out to be distressingly wrong.
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 12 of 13
1. Hollstein, E. Mitteleuropäische Eichenchronologie. Verlag Philipp von Zabern, Mainz am Rhein
2. Pilcher, J.R., Baillie, M.G.L., Schmidt, B. and Becker, B., 1984. A 7,272-year tree-ring chronology
for western Europe. Nature 312 (5990), 150-152.
3. Radiocarbon (1986), Vol.28, No.2A/B.
4. Ossowski Larsson P. & Larsson L.Å. (2015). Dendrochronological dating of Roman time.
ResearchGate DOI: 10.13140/RG.2.1.5129.6806.
5. Schmidt B., Dating of Roman sites. In: Dendrochronology and Archaeology in Europe: Proceedings
of a Workshop of the European Science Foundation (ESF), held in Hamburg, April 28-30, 1982
Vol.141 of Bundesforschungsanstalt für Forst- und Holzwirtschaft Hamburg: Mitteilungen der
Bundesforschungsanstalt für Forst- und Holzwirtschaft, Hamburg
6. K. Nicolussi, M. Kaufmann, Thomas M. Melvin, J. van der Plicht, P. Schießling, and A. Thurner
(2009). A 9111 year long conifer tree-ring chronology for the European Alps: a base for environmental
and climatic investigations. The Holocene 19(6):909-920.
7. Nicolussi, K., Lumassegger, G., Patzelt, G., Pindur, P. and Schießling, P. (2004). Aufbau einer
holozänen Hochlagen-Jahrring-Chronologie für die zentralen Ostalpen: Möglichkeiten und erste
Ergebnisse. In Innsbrucker Geographische Gesellschaft, editor, Innsbrucker Jahresbericht 2001/2002
16, 114–136. https://www.uibk.ac.at/geographie/forschung/dendro/publikationen---pdf-files/2004-
8. Becker, B. (1981). Fällungsdaten Römischer Bauhölzer anhand einer 2350-jährigen Süddeutschen
Eichen-Jahrringchronologie. Fundberichte aus Baden Württemberg No.6, 369-386.
9. Büntgen U., Tegel W., Nicolussi K., McCormick M., Frank D., Trouet V., Kaplan J.O., Herzig F.,
Heussner K.U., Wanner H., Luterbacher J., Esper J. (2011). 2500 Years of European Climate
Variability and Human Susceptibility, Science 331, 578-582.
10. Stuiver, M. (1982). A high-Precision calibration of the AD radiocarbon time scale: Radiocarbon,
24(1), 1-26. https://journals.uair.arizona.edu/index.php/radiocarbon/article/view/740/745
11. Michael McCormick, Ulf Büntgen, Mark A. Cane, Edward R. Cook, Kyle Harper, Peter Huybers,
Thomas Litt, Sturt W. Manning, Paul Andrew Mayewski, Alexander F. M. More, Kurt Nicolussi, Willy
Tegel (2012). Climate change during and after the Roman empire: Reconstructing the past from
scientific and historical evidence. Journal of Interdisciplinary History 43, 169–220.
12. PAGES 2k Consortium (2013). Continental-scale temperature variability during the past two
millennia. Nature Geoscience 6, 339-346.
Possible errors in the Eastern Alpine Conifer Chronology, draft, 2017-02-02, Page 13 of 13
13. Glur L, Wirth SB, Büntgen U, Gilli A, Haug GH, Schär C, Anselmetti FS (2013). Frequent floods in
the European Alps coincide with cooler periods of the past 2500 years. Nature – Scientific Reports 3:
2770; doi: 10.1038/srep02770 http://buentgen.com/wp-
14. 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 & T. E. Woodruff (2015). Timing and climate forcing of volcanic eruptions for the past 2,500
years. Nature 523, 543–549. doi:10.1038/nature14565.
15. Ulf Büntgen, Vladimir S. Myglan, Fredrik Charpentier Ljungqvist, Michael McCormick, Nicola Di
Cosmo, Michael Sigl, Johann Jungclaus, Sebastian Wagner, Paul J. Krusic, Jan Esper, Jed O.
Kaplan, Michiel A. C. de Vaan, Jürg Luterbacher, Lukas Wacker, Willy Tegel and Alexander V.
Kirdyanov (2016). Cooling and societal change during the Late Antique Little Ice Age from 536 to
around 660 AD. Nature Geoscience 9(3). DOI: 10.1038/ngeo2652.
16. Luterbacher J, Werner J, Smerdon J, Barriopedro D, Fernández-Donado L, Gonzalez-Rouco JF,
Barriopedro D, Ljungqvist F, Büntgen U, Zorita E, Wagner S, Esper J, McCarroll D, Toreti A, Frank D,
Jungclaus J, Barriendos M, Bertolin C, Bothe O, Brázdil R, Camuffo D, Dobrovolný P, Gagen M,
García-Bustamante E, Ge Q, Gómez-Navarro J, Guiot J, Hao Z, Hegerl G, Holmgren K, Klimenko V,
Martín-Chivelet J, Pfister C, Roberts N, Schindler A, Schurer A, Solomina O, von Gunten L, Wahl E,
Wanner H, Wetter O, Xoplaki E, Yuan N, Zanchetti D, Zhang H, Zerefos C (2016). European summer
temperatures since Roman times. Environmental Research Letters 11: 024001, doi: 10.1088/1748-
17. Esper J, Büntgen U, Timonen M, Frank DC (2012) Variability and extremes of northern
Scandinavian summer temperatures over the past two millennia. Global and Planetary Change 88-89:
18. R E Taylor, W C Beaumont, J Southon, D Stronach, D Pickworth (2010). Alternative Explanations
for Anomalous 14C Ages on Human Skeletons Associated with the 612 BCE Destruction of Nineveh.
Radiocarbon, Vol 52 (2), p 372-382.