While writing about the recent purported discovery of Noah’s Ark, supposedly dated to about 4800 years old, I was reminded of an article I read several weeks ago. It has to do with carbon-14 (C-14 or 14C) dating and its limitations.
Without getting into too much detail, trace amounts of carbon-14 are found in atmospheric carbon dioxide (CO2). While carbon-14 “beta decays” into nitrogen-14, it is constantly being replaced from cosmic rays bombarding the nitrogen-14 and turning it back into carbon-14. So, the amount of carbon-14 in the atmosphere is assumed to be roughly constant. (More on this later.)
Carbon-14 dating, or radiocarbon dating, is rather unique among radiometric dating methods, in that it can only date organic matter — i.e., things that used to be alive. While a plant is alive, it absorbs and fixes (i.e., converts into a solid) carbon-14 from the CO2 it takes in via photosynthesis. Animals, on the other hand, absorb carbon-14 when they eat plants and/or other animals. Once a living thing dies, it doesn’t take in anymore carbon-14. The percentage of carbon-14 in the remains, which had been assumed to be equal to that in its environment while alive, begins to decrease as it decays back into the non-radioactive isotope nitrogen-14. So, the amount of carbon-14 left behind is used to determine how long ago it died.
Radiogenic dating methods are typically good only for dating things that fall in the range of about six to seven times the length of the parent isotope’s half-life. If it’s any older than that, there is just not enough left of the element to get a decent sample for measuring accurately. The half-life for carbon-14 is about 5730 years, so radiocarbon dating is not usually attempted for objects believed to be over about 40,000 years. Or, at least, the accuracy is understood to be less reliable. (Though some have placed the technical limit at 60,000 +/- 2,000 years.) The ability to not only directly detect but count the decay of individual C-14 atoms during analysis has greatly increased since the advent of Accelerator Mass Spectrometry. Use of AMS technology, along with improved calibration methods (see below), have begun pushing the practical limits into the 45,000 to 50,000 years range.
Even with a good sample and following proper protocol, the “raw” radiocarbon date is still an approximation (i.e., error bars are too big to be very useful, in many cases). This is because of various factors, both natural and unnatural, that can affect the initial amount of carbon-14 in the atmosphere. Natural factors include the altitude where the sample was found, local volcanic eruptions, huge carbon reservoirs, the Earth’s magnetic field, fluctuations in solar activity, among others. Unnatural factors include things like heavy local industrialization and above-ground nuclear testing (since 1950). So, scientists have come up with ways to compensate. They look at other, independent dating methods/sources that are known to fall in the same range, sometimes with some established dates based on historical events, and compare those date estimates. The data from these are used to construct what is called a ‘calibration curve’. The curve is then applied to adjust the “raw” date, giving a much more accurate age for the object in question.
New developments have resulted in the publication of probably the most accurate radiocarbon calibration curve, yet. It’s referred to as the INTCAL09 standard.
For the past 30 years, a group called IntCal (International Calibration?) has been steadily building and refining a superior calibration curve. They began with dendrochronology — i.e., tree-ring dating. Records from thousands of overlapping tree-ring segments from the Northern Hemisphere were used, which provide a very accurate check of raw radiocarbon dates and how much they must be corrected. [Please note that dendrochronologies are built from many overlapping specimens, not from single trees.] But the oldest trees can only provide good, calibrated data to about 12,400 years.
To take it to the next stage, geochronologist Paula Reimer and her team had to use several other sources that are not quite as precise — e.g., corals and fossilized foraminifers (single-celled organisms that secrete calcium carbonate). This got the curve to about 26,000 years, where the foraminifer and coral data stayed in pretty close agreement. Beyond that, however, the data sets diverged significantly — up to several thousands of years — and the group could not agree on how to handle the differences. So, they published their efforts to that point (2004), but there seemed to be no hope in progressing to the hoped-for 50,000 years goal.
Then, not long ago, the IntCal group got access to new and more accurate data from foraminifers, corals, and other sources. They were also able to apply some rather sophisticated statistical algorithms to help determine which way data gaps bend the curve. Thus, they were able to resolve most of the problems and come to a consensus. The resulting INTCAL09 was published in Radiocarbon this past January (2010). Not only does it now extend the calibration curve out to 50,000 years, but earlier sections were able to be improved upon, as well.
Archaeologists and anthropologists are excited about the extended radiocarbon dating ability. It should help them in their efforts to track the cultural development and migrations of early “modern” humans, which are usually left to imprecise dating methods like thermoluminescence. For example, the new calibration curve has already shown that the earliest paintings at Chauvet Cave in southern France are 36,500 years old (during a period of relative warmth) rather than the previous estimate of 32,000 years (directly following a major cold spell).
But, wait! There’s more!
The latest findings about “the Earth’s carbon reservoirs and how they changed over time” are anticipated to be factored into an updated IntCal curve in 2011, says Peimer. Such fun!
* Thanks to atronomer/author David Darling of The Internet Encyclopedia of Science for use of the C-14 image.