2 Radiocarbon Dating and Calibration
Learn how radiocarbon measurements are made, why samples and calibration choices matter, and how to interpret and report calibrated date ranges with appropriate caution.
What measures
estimates when material from a once-living organism stopped exchanging carbon with its environment. It dates the organism’s death; it does not automatically date the making of an artifact, the formation of a layer, or another event associated with the sample. A sound interpretation depends on how closely the sample relates to the event in question and how the laboratory result is calibrated.
From carbon uptake to a
Cosmic rays help produce carbon-14 in the atmosphere. Living organisms take in carbon, including a small amount of carbon-14, through processes such as photosynthesis and feeding. After death, an organism no longer replenishes its carbon-14, which decays at a known rate. Measuring the remaining carbon-14 relative to stable carbon isotopes gives a .
An uncalibrated result is conventionally reported in radiocarbon years , where “present” means AD 1950. For example, is a measured with uncertainty, not a calendar date of 50 BC. The age calculation follows standardized conventions, including a correction for isotopic fractionation; it does not assume that atmospheric carbon-14 was constant through time.
Choose and prepare a suitable sample
Choose a sample whose relationship to the question is clear. Short-lived plant remains from a well-defined context may closely date the event of interest. By contrast, the means a piece of long-lived wood can predate its burning or reuse by decades or centuries. Even genuinely ancient material may be poorly related to the event being dated if it was moved, reused, or redeposited.
The laboratory needs suitable material in sufficient quantity. Selection should consider preservation, context, possible mixing, and contamination from soil, handling, storage, or conservation treatments. Laboratories physically and chemically pretreat samples to remove unwanted material, using methods suited to the sample type. Pretreatment cannot correct a mistaken association between sample and event, and unsuitable or contaminated samples may produce misleading results. Consult the dating laboratory about material choice and handling.
How laboratories measure carbon-14
Laboratories commonly use one of two approaches:
counts carbon isotopes directly and can measure small samples.
Radiometric methods measure radiation produced as carbon-14 decays and generally require more material.
A laboratory report gives an identifier, a conventional or equivalent carbon-14 measurement, and an analytical uncertainty. That uncertainty describes measurement precision. It does not cover every possible source of error, including contamination, an inappropriate sample, or an unrecognized .
Convert measurements to calendar dates
Atmospheric carbon-14 has varied over time, so a cannot be converted to a calendar date by simply subtracting it from 1950. compares the measured result with a built from samples with independently known calendar ages, including dated tree rings. It maps the measurement and its uncertainty to a probability distribution of possible calendar ages.
Use a curve suited to the carbon source. Terrestrial Northern Hemisphere samples are generally calibrated with IntCal20, Southern Hemisphere samples with SHCal20, and marine samples with Marine20, with an appropriate local reservoir correction when needed. Freshwater and mixed-diet samples may also need special consideration because their carbon can come from sources with different apparent radiocarbon ages.
Interpret calibrated date ranges
commonly produces one or more calendar-date ranges rather than a single exact year. The shape of the matters: flat or irregular sections can yield broad ranges or several separated ranges, even when the laboratory measurement is precise. Greater measurement precision therefore does not always produce a narrowly resolved calendar date.
A calibrated result is a probability statement. A report giving a range at probability means the calibrated probability distribution assigns of its probability to the reported range or ranges under the stated model and assumptions. It does not mean that every year within a range is equally likely. If yields multiple ranges, report them rather than selecting one without justification.
Report dates and recognize limitations
Keep uncalibrated distinct from calibrated calendar-year notations such as cal , cal BC, and cal AD. A clear report identifies the sample and laboratory, the and analytical uncertainty, the and software or method, the probability level, and any reservoir correction.
Contamination can make a result appear too young or too old. Old wood and reused material can date earlier than the event of interest, while reservoir effects can make organisms appear older than contemporaneous terrestrial organisms. -curve plateaus can limit calendar precision. is useful for suitable carbon-bearing material from roughly the recent past to about 50,000 years ago, but results near its effective limit require particular caution. Treat a radiocarbon date as evidence to combine with archaeological context and other chronological information, not as a stand-alone date for an entire site or event.