Getting the noise out of the data

Kiel-based research team helps to redetermine the radiocarbon calibration curve


Radiocarbon dating is the standard procedure used to determine the age of, for example, archaeological finds such as bones. Even past climate events can be dated in this way. In the case of particularly old samples, previous results were often inaccurate and there were large temporal discrepancies, sometimes of several thousand years. An international project involving researchers from Kiel University (CAU) has recalculated the calibration curves for radiocarbon dating in the Northern and Southern Hemisphere and in the ocean and significantly reduced the temporal fluctuations. The researchers have published their findings in the current edition of the specialist journal Radiocarbon.

Carbon isotope 14C is generated in the atmosphere by cosmic radiation. Plants absorb it through photosynthesis, humans and animals through the food chain. As the isotope’s half-life of 5,730 years can be determined relatively accurately and the decay process runs uniformly, age can be determined very well by determining the 14C content. It works for finds of up to 60,000 years old. The 14C content in the atmosphere has not always been at the same level and therefore its content in organisms has also varied. Researchers are continuously working on removing this confounding variable and determining the radiocarbon calibration curve more accurately. The calibration curve is used as a reference for exact age determination. If, for example, the 14C value of a sample is dated to 5300 BC, this corresponds to the linear half-life of the isotope. In order to incorporate atmospheric influences into the dating process, the researchers compare this value to the calibration curve and by doing so they can assign the sample to a precise period in the history of the Earth.

Kiel-based researchers provided data on the ocean surface

Over a period of seven years, a research group of 42 researchers from 39 institutes in 14 countries have assessed more than 15,000 radiocarbon calibration measurements dating up to 60,000 years old. From this they created new, much more accurate calibration curves for the northern and southern hemisphere (IntCal20 and SHCal20) and for the oceans (Marine20). In doing so, they combined data from tree rings, stalagmites, corals and sediment cores from seas and oceans using new statistical methods. "Technical advances and a much smaller required sample quantity considerably simplified the sample taking," said Professor Pieter M. Grootes of the Johanna Mestorf Academy and the Institute for Ecosystem Research at the CAU. For example, earlier samples from tree trunks had to have ten to 20 annual growth rings to provide enough wood, i.e. carbon. The latest technology requires a thousand times less material for the evaluation: it can be dated from just one annual growth ring. “For the calibration curves we still prefer trees that are hundreds of years old, so that we can use the number of annual growth rings as calibration for the 14C measurements. In this way, so far, we have already been able to go back up to 14,000 years,” added Grootes. For the period before that, he continued, we have to switch over to other independently datable archives, such as the above-mentioned stalagmites, corals or oceanic sediment cores.

Pieter M. Grootes and Professor Michael Sarnthein of the Institute of Geosciences, also at Kiel University, have contributed and evaluated data on the 14C content in the paleo-ocean – i.e. oceanic bodies of water considered in terms of the history of the Earth. They have thereby helped to improve our understanding of environmental influences on the oceanic 14C content and thus enhance incorporation of the oceanic 14C archives in the atmospheric calibration curves. “We are getting the noise out of the data,” said Grootes, describing the basic principle of the new results. “The previous calibration curves from 2013 treated the faster fluctuations as noise and were therefore heavily smoothed. As a result, important information was lost regarding decadal fluctuations in atmospheric 14C content and solar events that allow for calibration that is exact to the year. In general, therefore, calibration was less accurate.” The new statistical processing of calibration data separates the short- and longer-term natural fluctuations in atmospheric 14C content from statistical “noise” of 14C measurement results. “We now have a much better understanding of how the ocean surface behaves in relation to the atmosphere. Previously the curves were sometimes based on assumptions, but now we have increasingly sound data,” explained Pieter M. Grootes.

“The redetermination helps us to put temporal sequences in order, for example. In the Collaborative Research Centre 1266 Scales of Transformation, we are investigating a shutdown around 3100 BC. If we can now place crisis events accurately to within ten to twenty years, it is possible for the causes to be better determined and differentiated on a temporal basis,” said Johannes Müller, Professor at the Institute of Prehistoric and Protohistoric Archaeology at Kiel University and spokesman for the CRC 1266, explaining the advantages of redetermination.

Original publications:

Original publications:
Reimer, P., Austin, W., Bard, E., Bayliss, A., Blackwell, P., Bronk Ramsey, C., ... Talamo, S. (2020). The IntCal20 Northern Hemisphere Radiocarbon Age Calibration Curve (0–55 CAL kBP). Radiocarbon, 1-33. DOI-Nummer:10.1017/RDC.2020.41.

Heaton, T., Köhler, P., Butzin, M., Bard, E., Reimer, R., Austin, W., … Skinner, L. (2020). Marine20 – The Marine Radiocarbon Age Calibration Curve (0–55,000 CAL BP). Radiocarbon, 1-42. DOI-Nummer:10.1017/RDC.2020.68.

Hogg, A., Heaton, T., Hua, Q., Palmer, J., Turney, C., Southon, J., .... Wacker, L. (2020). SHCal20 Southern Hemisphere Calibration, 0–55,000 Years CAL BP. Radiocarbon, 1-20. DOI-Nummer:10.1017/RDC.2020.59.



Prof. Dr. Pieter Grootes
Johanna Mestorf Academy
Leibniz Labor
 +49 (0)431 880-1229


Figure with two graphs.
© Sahra Talamo

Overlap between the existence of Neanderthals and Home Sapiens: In comparison, the seven-year-old curve IntCal13 and the new curve IntCal20.

Close-up of pine tree-rings.
©P. Brewer, University of Arizona

Bristlecone pine tree-rings from the Thera period (second millennium BC). Multiple replicated measurements of radiocarbon in single rings from ancient trees like these now cover the period 1700-1500 BC in IntCal20.

Stone tomb on a meadow.
© Sara Jagiolla, Institut UFG

The improvement of the 14C calibration curve has a direct impact on the accuracy of the dating of archaeological features. Among other things, construction phases of megalithic tombs - such as here the Brutkamp in Albersdorf - whose history can now be studied in the Kiel Collaborative Research Centre "TransformationsDimensions of Prehistoric and Archaic Societies", can now be dated much better. For example to the year 3360 BC plus/minus ten years.

Photo of analytical equipment.
© Ionplus

View inside the ion source of an AMS system. Samples to be analysed are

introduced in the measurement position from the left hand side. The gloving part is the

ioniser that produces a cesium ion which decompose the sample material to produce an ion

beam of the different carbon isotopes 12C, 13C and 14C.

Christin Beeck
Press, Digital and Science Communication