The the latest calibration curve is here:

http://www.radiocarbon.org/IntCal13%20files/intcal13.pdf

You can see that there are some big ambiguities over the last ~300 years ago, but if you have other evidence to the date your samples (for instance unburnt wood with tree rings)  then you could use those curves to narrow the dates down.

Thank you for your reply, Alastair, and for the link.

My understanding is that inclusion of additional evidence for calibration, like wood from tree rings, would require C14-dating of the rings themselves?

Dear Aaron,

you don't need to date tree rings by 14C, you can date them directly by dendrochronology. You have to make sure if a (dendrochronological) master chronology is available for your region. This is the critical point, as master chronologies are not available for many geographic areas and for all species of plants.

Best, Diego

Dear Diego,

Thanks again for your reply, I actually have a detailed master chronology for the region obtained from the NOAA website (https://www.ncdc.noaa.gov/data-access/paleoclimatology-data/datasets/tree-ring) with ring widths, etc.

I was hoping to use dendrochronology and the master chronology to date the oldest trees at my sites, given the uncertainty surrounding the C14 dates (suess effect) but am worried that some of the fires may pre-date the oldest trees at my sites (and make it difficult with dendro). The master chronology extends back as far as 1822. However, while collecting charcoal samples for C14 dating, I also collected several cores at each site. I have several cores as old as 240 years, thereby allowing me to extend the master chronology a bit further, I think (sampled in 2013, so ~ 1770 ish).

Is it possible to somehow correct for uncertainty in carbon-dating estimates with tree ages?

Hi Aaron,

The Suess effect does not produce an uncertainty. It gives you a set of values rather than just one. OTOH, as Diego is saying tree rings give you one exact age. C14 ages are calibrated using tree rings!

Let's assuming your 14C age for one of your samples is 150 years BP then because of the Suess effect the chart  I cited gives you four ages:  50, 150, 220 and 270 BP cal.  If the age from your tree rings for the same sample corresponds with only one of these dates then that is very likely the correct one.

So, you can work backwards on that chart and use your  cal. tree ring age to find the corresponding  C14 age, and check that it agrees with the C14 age from the lab., which I guess is what you were wanting to do anyway.

HTH.

Hello Aaron,

it’s a problem I’m, alas, all too familiar with. There is no fool proof way to date anything by 14C after the middle of the 17th century. If your charcoal samples are large enough, you might try wiggle-matching, but even then, the fact that there are several overlapping peaks/troughs in the calibration curve make that a very iffy endeavour. As for dendrochronology: you will always need series of more than 30 rings, preferably over 50, to securely lock them on to your master chronology. And it still is a statistical probability you get for a match. You can even try and use charcoal for dendrochronology, but it’s quite hard, as the rings have been warped by the carbonisation, so you will need a minimum of 50 rings to get a half way reliable date. And that is something very hard to realise when working with the small particles you normally get from soils.

Good luck,

Rengert

Thank you all for your insightful comments. It seems clear that some inclusion of tree rings will be necessary to date my samples. I guess this leads me to ask - is the C14 data useful beyond confirming fire-dating methods like dendro?

For instance, I played around a bit with the Calib program available online here (http://calib.qub.ac.uk/calib/) which uses IntCal13 but am having trouble interpreting the estimate. It asks for both 'radiocarbon age' and 'age uncertainty' and gives me multiple values corresponding to time periods with what I believe are probability values. Is this what you meant about multiple sets of values, Alastair?

My main concern is that  fire may pre-date the oldest trees at some of my sites. I'm hoping that with the master chronology I can also date dead trees that have died from fire by matching up the rings, or by finding fire scars, although my initial visits have failed to find any.

Thanks again,

Aaron

Dear Aaron,

14C datings are reported in either of two forms, which always start with the laboratory-identification code like OxA (Oxford), GrA (Groningen), MAMS (Mannheim), just to name a few here from Europe, followed by the lab-number. Then you either get a date followed by its confidence-interval or the percentage modern radiocarbon with its standard error. Here two real life examples from a paper that we published last month (see below).

J-8544 842±23 BP 90.05 0.26

J-8543 665±24 BP 92.06 0.27

Now if you feed these data into either CALIB or OxCAL (I happen to use the latter) you get the calibrated data. Now look on the graphic representation (with Oxcal there is the dropdown menu showing table, where you choose ‘single plot’). With the first value you see the probability-distribution in red on the vertical axis, which is then ‘projected’ on the calibration curve, which results in the grey distribution on the horizontal axis. This procedure results in the calibrated date, which lies with 95.4% probability between 1161 and 1255 AD.

Now take the second date and feed it into the program. In this part of the curve there is a nasty wiggle which causes the distribution to hit the curve twice, giving two well-separated probability intervals. Your 95.4% probability is now split nearly equally between the two date-ranges of 1278-1316 and 1355-1390 AD. So you can be pretty sure your piece of charcoal doesn’t date to any period outside these two intervals (including the range of 1317-1354), but there is absolutely no way to decide in which of the two humps the actual date lies. And this is exactly the effect you get with any date younger than ca. 270 BP, which is spread out over at least three intervals, as the calibration curve (which corresponds to the concentration of 14C in the air), starts to yo-yo very badly after ca. 1520.

Hope this clarifies the procedure a bit and shows why 14C after the (European) Middle Ages aren’t of much use.

Greetings,

Rengert

Article Forest exploitation for charcoal production and timber since...

That clarifies it! Thank you Rengert!