This module expects one or more rows of radiocarbon dates, with two columns for uncalibrated age (BP 1950) and its laboratory-reported standard error (one-sigma). Two calibration curves are included, IntCal20 (Reimer et al. 2020) for atmospheric (terrestrial) samples and MarineCal20 (Heaton et al. 2020) for marine samples. Please include a reference to Reimer et al. (2020) or Heaton et al. (2020) when reporting calibration results (as well as Past of course!).
For the marine calibration, a globally averaged reservoir age (around 400 years, but varying with time) is included in the calibration. However, a local reservoir age correction (delta R) and its one-sigma error should be entered. These values can be looked up e.g. in the Marine Reservoir Correction Database at http://calib.org/marine/
The IntCal20 curve is inaccurate in the Southern Hemisphere. The Southern Hemisphere calibration curve (SHCal20) will be included in a future version.
The output ages can be reported in cal BP 1950 (years before 1950), BP 2000, or CE (Common Era). The CE values are reported with negative or positive sign, and with a year zero included (ISO 8601).
Plotted probability curves are normalized to a peak value of 1.
Calibrated ages should always be reported with confidence intervals (and ideally the complete probability curve), but the median and mode (age of peak probability) are also given by Past as convenient short-hands. The median is more commonly used, but Michczynski (2007) found the mode to perform better in simulation studies.
Confidence intervals
For consistency with other modules in Past, the confidence intervals on calibrated ages are given with respect to the 95% range, not 95.4% as in some other calibration software. This can give slightly shorter confidence intervals in Past.
The “Percentile” option calculates a single confidence interval from the 2.5 to the 97.5 percentile points. The “Level set” option calculates a probability threshold such that the sum of all probabilities larger than the threshold is 0.95. All the corresponding t values are included in the CI. Because the probability curve can be multimodal, this can give rise to several disjoint segments in the CI. Past reports up to three of these segments containing the largest areas (reported in %), which is usually sufficient to reach 95%.
Both these options produce valid 95% confidence intervals. The “Level set” option gives a more detailed view, and is used in most other calibration software, but the “Percentile” option is more convenient and gives a sufficiently informative CI for most purposes (personal opinion!). If only a single segment is sufficient for the level set CI, this will normally be very similar to the percentile CI.
The confidence interval is shown as red lines in the plot.
For computational details, see the Past manual.
References
Heaton TJ, Köhler P, Butzin M, Bard E, Reimer RW, Austin WEN, Bronk Ramsey C, Grootes PM, Hughen KA, Kromer B, Reimer PJ, Adkins J, Burke A, Cook MS, Olsen J, Skinner LC. 2020. Marine20-the marine radiocarbon age calibration curve (0-55,000 cal BP). Radiocarbon 62. doi: 10.1017/RDC.2020.68.
Michczynski, A. 2007. Is it possible to find a good point estimate of a calibrated radiocarbon date? Radiocarbon 49:393-401.
Reimer P, Austin WEN, Bard E, Bayliss A, Blackwell PG, Bronk Ramsey C, Butzin M, Cheng H, Edwards RL, Friedrich M, Grootes PM, Guilderson TP, Hajdas I, Heaton TJ, Hogg AG, Hughen KA, Kromer B, Manning SW, Muscheler R, Palmer JG, Pearson C, van der Plicht J, Reimer RW, Richards DA, Scott EM, Southon JR, Turney CSM, Wacker L, Adolphi F, Büntgen U, Capano M, Fahrni S, Fogtmann-Schulz A, Friedrich R, Köhler P, Kudsk S, Miyake F, Olsen J, Reinig F, Sakamoto M, Sookdeo A, Talamo S. 2020. The IntCal20 Northern Hemisphere radiocarbon age calibration curve (0-55 cal kBP). Radiocarbon 62. doi: 10.1017/RDC.2020.41.