IntCal 2020

14 Aug 2020

At https://phys.org/news/2020-08-secrets-international-carbon-dating-standa... ... (also at www.heritagedaily.com/2020/08/researchers-unlock-secrets-of-the-past-wit...) ... a team of researchers from the universities of Sheffield, Belfast, Bristol, Glasgow and St Andrews, in collaboration with international colleagues, have come up with a revision of the C14 dating system, IntCal 2020. It used measurements from around 15,000 samples from a variety of objects as part of a seven year project. The claim is they can now date objects more accurately. C14 dating is expensive and can stretch project funds. The methodology also has the problem of plateau events and calibration is designed to iron these out. Nowadays, there are a number of rival dating methodologies so it is imperative for laboratories to have as much precision as possible. Whilst, no doubt, these calibration revisions are no problem for professionals, who are able to distinguish between them, it does present a source of confusion to amateurs, and those with a passing interest in archaeology. Dates change. New researchers, fresh out of university and conversant with the most up to date calibration, will confidently quote the latest date, not realising, perhaps, they are contradicting somebody, somewhat earlier in time, who has equally and with confidence, quoted an entirely different date. No doubt it doesn't matter when it comes down to small changes in time, reducing the uncertainty levels by a few decades. However, there are examples where re-calibration has caused major changes and has definitely created confusion amongst Joe Public. This was particularly  true of earlier calibrations which have tended to raise dates in the 3rd millennium BC. What had originally been dated roughly at around 2300BC, or a little earlier, at Stonehenge for example, was suddenly upped to near 2500BC. It didn't just effect the dating of objects, or re-erections and changes in the position of stones in the monument, but pushed back the arrival of the Beaker Folk. When this was combined with recent genetic research, what was once described as a Neolithic monument was transformed virtually into a Bronze Age project, as the Neolithic people had declined in importance as a result of newcomers from the near continent, bringing metal working skills with them. In fact, 2300BC was transformed. At SIS Review, for example, we had a series of articles in which the author wrote about geological, archaeological, and human movements that he broadly called the 2300BC event. It combined the tree ring low growth events at 2345BC and 2200BC so it was never strictly accurate, but the title is now completely misleading as these events have been pushed forward by a considerable number of years. Likewise, when consulting articles written 20 years ago or longer one can no longer accept the dates used by the authors. It may also apply to museum collections. I suppose that must depend on how big a museum might be and how much resources it has at its disposal, and how much it wants to keep abreadst of the dating methodologies. Basically, the new calibration is designed to aid archaeologists, and others, including museum curators, to get a more accurate date. Joe Public is not really part of the equation.

IntCal20 is more accurate because it has more data included in the curve, some portions of which are constructed from single year tree ring measurements. This increases the resolution of the curve, as well as certain other factors of the older curve. There is thus better confidence in the accuracy of a calibrated date with regards previous calibration curves. It also means that the precision of a calibrated date could increase, or decrease, over a previous iteration of the curve due to slight changes in the shape of the new curve. This might apply to the Thera eruption for example. Short lived samples from the volcano's destruction layer produce a date of around 3350 +/- 10 BP. In IntCal13 this calibrated to 1683-1672 and 1666-1617 cal BC (at 2 sigma confidence) but to the ranges of 1733-1719, 1688-1651, 1645-1608, 1604-1602, and 1583-1559 cal BC for IntCal20 (at 2 sigma). This portion of the curve is constructed from many contiguous single year radiocarbon measurements of tree rings and so is considered most accurate. Due to the change of the curve it has resulted in a decrease in precision of the calibration. Similarly, if the wiggle match calibration for the olive branch buried by the Thera eruption, a date age range for the outer ring of 1625-1608 cal BC for IntCal13 can be compared to 1617-1578 cal BC for IntCal20. Greater confidence in accuracy of the calibration but less precise. The decrease in precision has resulted from a shift of the curve meaning that the radiocarbon age falls upon a prolonged radiocarbon plateau, resulting in increased smear in the calibrated date. This does not mean IntCal20 has sacrificed precision for accuracy as when the calibrated date for the Laacher See eruption in Germany, the precision has increased. A 200 year uncertainty in IntCal13 has become a 160 year uncertainty in IntCal20. For more informatioon see https://www.cambridge.org/core/journals/radiocarbon/article/recent-sampl... ...