Calibration is not only done before an analysis but also on analytical results as in the case of radiocarbon dating—an analytical method that identifies the age of a material that once formed part of the biosphere by determining its carbon-14 content and tracing its age by its radioactive decay. Carbon-14 is a naturally occurring isotope of the element carbon. It is also called “radiocarbon” because it is unstable and radioactive relative to carbon-12 and carbon-13. Carbon consists of 99% carbon-12, 1% carbon-13, and about one part per million carbon-14. Results of carbon-14 dating are reported in radiocarbon years, and calibration is needed to convert radiocarbon years into calendar years.
Uncalibrated radiocarbon measurements are usually reported in years BP where 0 (zero) BP is defined as AD 1950. BP stands for “Before Present” or “Before Physics” as some would refer to it. It should be noted that a BP notation is also used in other dating techniques but is defined differently, as in the case of thermoluminescence dating wherein BP is defined as AD 1980. It is also worth noting that the half-life used in carbon dating calculations is 5568 years, the value worked out by chemist Willard Libby, and not the more accurate value of 5730 years, which is known as the Cambridge half-life. Although it is less accurate, the Libby half-life was retained to avoid inconsistencies or errors when comparing carbon-14 test results that were produced before and after the Cambridge half-life was derived. Radiocarbon measurements are based on the assumption that atmospheric carbon-14 concentration has remained constant as it was in 1950 and that the half-life of carbon-14 is 5568 years. Calibration of radiocarbon results is needed to account for changes in the atmospheric concentration of carbon-14 over time. These changes were brought about by several factors including, but not limited to, fluctuations in the earth’s geomagnetic moment, fossil fuel burning, and nuclear testing. The most popular and often used method for calibration is by dendrochronology.
The science of dendrochronology is based on the phenomenon that trees usually grow by the addition of rings, hence the name tree-ring dating. Dendrochronologists date events and variations in environments in the past by analyzing and comparing growth ring patterns of trees and aged wood. They can determine the exact calendar year each tree ring was formed. Dendrochronological findings played an important role in the early days of radiocarbon dating. Tree rings provided truly known-age material needed to check the accuracy of the carbon-14 dating method. During the late 1950s, several scientists (notably the Dutchman Hessel de Vries) were able to confirm the discrepancy between radiocarbon ages and calendar ages through results gathered from carbon dating rings of trees. The tree rings were dated through dendrochronology. At present, tree rings are still used to calibrate radiocarbon determinations. Libraries of tree rings of different calendar ages are now available to provide records extending back over the last 11,000 years. The trees often used as references are the bristlecone pine (Pinus aristata) found in the USA and waterlogged Oak (Quercus sp.) in Ireland and Germany. Radiocarbon dating laboratories have been known to use data from other species of trees.
In principle, the age of a certain carbonaceous sample can be easily determined by comparing its radiocarbon content to that of a tree ring with a known calendar age. If a sample has the same proportion of radiocarbon as that of the tree ring, it is safe to conclude that they are of the same age. In practice, tree-ring calibration is not as straightforward due to many factors, the most significant of which is that individual measurements made on the tree rings and the sample have limited precision so a range of possible calendar years is obtained. And indeed, results of calibration are often given as an age range rather than an absolute value. Age ranges are calculated either by the intercept method or the probability method, both of which need a calibration curve.
The first calibration curve for radiocarbon dating was based on a continuous tree-ring sequence stretching back to 8,000 years. This tree-ring sequence, established by Wesley Ferguson in the 1960s, aided Hans Suess to publish the first useful calibration curve. Suess’s curve, based on the bristlecone pine, used tree rings for its calendar axis. There have been many calibration curves published since Suess’s curve, but their proliferation brought more problems than solutions. In later years, the use of accelerator mass spectrometers and the introduction of high-precision carbon dating have also generated calibration curves. A high-precision radiocarbon calibration curve published by a laboratory in Belfast, Northern Ireland, used dendrochronology data based on the Irish oak. Nowadays, the internationally agreed upon calendar calibration curves reach as far back as about 48000 BC (Reimer et. al., INTCAL13 and Marine13 radiocarbon age calibration curves 0 – 50000 yrs cal BP, Radiocarbon 55(4), 2013). For the period after 1950, a great deal of data on atmospheric radiocarbon concentration is available. This post-modern data is very useful in some cases in illustrating a calendar age of very young materials (Hua, et. al. Atmospheric Radiocarbon for the period 1950-2010, Radiocarbon, 55(4), 2013). A typical carbon-14 calibration curve would have a calendar or dendro timescale on the x-axis (calendar years) and radiocarbon years reflected on the y-axis.
The use of cal BC, cal AD, or even cal BP is the recommended convention for citing dendrochronologically calibrated radiocarbon dating results. Carbon dating results must be clear, hence they should not be reported simply as BC, AD, or BP. Cal BC and cal AD correspond exactly to normal historical years BC and AD, while cal BP refers to the number of years before 1950.
Carbon dating results must include the uncalibrated results, the calibration curve used, the calibration method employed, and any corrections made to the original result before calibration. The confidence level corresponding to calibrated ranges must also be included.
Last Updated May 5, 2016