|Landscape in Carina Nebula [Courtesy NASA]
How reliable is radiocarbon dating?
David H. Bailey
Updated 31 March 2019 (c) 2019
Radiocarbon dating, which is also known as carbon-14 dating, is one widely used radiometric dating scheme to determine dates of ancient artifacts. In discussions of the age of the Earth and the antiquity of the human race, creationists often assail perceived weaknesses in radiocarbon dating. Henry M. Morris, for instance, wrote, "Despite its high popularity, [radiocarbon dating] involves a number of doubtful assumptions, some of which are sufficiently serious to make its results for all ages exceeding about 2000 or 3000 years, in serious need of revision." [Morris2000, pg. 162].
How does radiocarbon dating work?
Radiocarbon dating is based on the fact that the interaction of cosmic rays from outer space with nitrogen atoms in the atmosphere produces an unstable isotope of carbon, namely radiocarbon. Since it is chemically indistinguishable from the stable isotopes of carbon (carbon-12 and carbon-13), radiocarbon is taken by plants during photosynthesis and then ingested by animals regularly throughout their lifetimes. When a plant or animal organism dies, however, the exchange of radiocarbon from the atmosphere and the biosphere stops, and the amount of radiocarbon gradually decreases, with a half-life of approximately 5730 years. Because of this relatively short half-life, radiocarbon is useful for dating items of a relatively recent vintage, as far back as roughly 50,000 years before the present epoch. Radiocarbon dating cannot be used for older specimens, because so little carbon-14 remains in samples that it cannot be reliably measured.
Creationists often criticize radiocarbon dating in the context of discussions of the age of the Earth. But, as is clear even from the very brief discussion in the previous paragraph, radiocarbon dating can say nothing one way or the other about whether the Earth is many millions of years old, since such dates are far beyond this method's range of resolution. Thus creationists and others who invoke perceived weaknesses in radiocarbon dating as justification to cast doubt on the great age of the Earth are either uniformed on very basic scientific facts, or else are highly being disingenuous to their audience.
Reliability of radiocarbon dating
Radiocarbon dating has been studied at great length over the past few decades, and its strengths and weaknesses are very well understood at this point in time. For instance, even in the 1950s, when Willard Libby first developed the process, it was recognized that the scheme assumes that the level of carbon-14 in the atmosphere is constant. But researchers have known at least since 1969 that the carbon-14 level has not been constant, so that the radiocarbon clock needs to be "calibrated."
As a result, various schemes are used to correct and calibrate radiocarbon dates, including:
In each case, radiocarbon dates, determined by well-established procedures and calculations, are compared directly with dates determined by the above methods, thus permitting the radiocarbon dates to be accurately calibrated with distinct and independent dating techniques.
- Dendochronology: counting tree rings.
- Measurements of coral or other carbonate structures such as stalagmites, corroborated using uranium-thorium radiometric dating.
- Optically stimulated luminescence dating. This is based on the fact that stimulating mineral samples with blue, green or infared light causes a luminescent signal to be emitted, stemming from electron energy that is proportional to the amount of background radiation the specimen has undergone since burial. This scheme can be used to date items between about 300 years to over 100,000 years, and thus can be used to double-check and calibrate radiocarbon dates [Optical2011].
- Varve sediments: Counting the alternating light and dark bands in glacial lake beds that record the annual passage of seasons.
In 2009, several leading researchers in the field established a detailed calibration of radiocarbon dating, based on a careful analysis of pristine corals, ranging back to approximately 50,000 years before the present epoch [Reimer2009]. Here is a graph showing radiocarbon dates on the vertical axis and the calibrated age on the horizontal axis (shown here with permission from Johannes van der Plicht, one of the authors of the 2009 study). The relative width of the red calibration curve indicates the range of uncertainty:
In October 2012, a team led by Christopher Ramsey of Oxford University published a new study, based on analyses of varves (alternating light/dark bands in sediments) from Lake Suigetsu, which is located about 350 kilometers west of Tokyo, near the coast of the Sea of Japan. These researchers collected core samples 70 meters deep, and then painstakingly counted the layers, year by year, to obtain a direct record stretching back 52,000 years. Comparing these counts with a series of 651 radiocarbon-dated samples spanning this record, they obtained a calibration curve that is very close to the 2009 calibration shown above [Callaway2012]. Thus these calibrations are very reliable indeed.
It should be emphasized that the actual calibrated dates are about 10%-20% older than the raw uncorrected radiocarbon dates that were once used. Compare, for example, the uncorrected line (blue dotted line) with the calibration curve (red curve). In other words, those hoping that uncertainties in radiocarbon dating, say in the assumption of constancy of atmospheric carbon-14 levels, will mean that specimens are really much younger than the measured dates, are in for a big disappointment -- it is now clear that specimens are actually somewhat older than the raw, uncalibrated reckonings.
Creationist criticisms of radiocarbon dating
As mentioned above, young-earth creationist writers have cited various anomalies and potential difficulties with radiocarbon dating, and have used these examples to justify their conclusion that the entire scheme is flawed and unreliable. For instance, creationist Walt Brown has pointed out inconsistencies in some radiocarbon dates of mammoths -- one part was dated to 40,000 years, another to 26,000 years (and wood surrounding it to 10,000 years), and yet another to between 15,000 and 21,000 years before the present epoch [Brown2001]. However, in the scientific results mentioned by Brown, the dates come from different mammoth specimens. Also, at least one of these dates comes from a hide that had been soaked in glycerin, rendering the date invalid. These and numerous other claimed anomalies in radiocarbon dating are explained in detail in Mark Isaak's book [Isaak2007, pg. 146-151].
In short, while like any other method of scientific investigation, radiocarbon dating is subject to anomalies and misuse, when used correctly in accordance with well-established procedures and calibration schemes, the method is a very reliable means of dating relatively "recent" artifacts. In any event, it must be emphasized once again that radiocarbon dating has no relevance one way or the other for the overall question of whether the Earth is many millions of years old, since the scheme can only be used to reliably date specimens less than approximately 50,000 years old.
Additional background is available in a well-written Wikipedia article on the topic [Radiocarbon2011],
and in Richard Wiens' article
See also, on this website, articles on the ages of the geologic periods
the reliability of radiometric methods
a "time machine" for studying the distant past
and the "uniformitarian" assumption and how it relates to evolution and the age of the Earth