quote:
I've got a question for you guys that seems to fit into this thread pretty well. Rei posted some specific events in which carbon dating should not be used, specifically volcanoes. My question is how do you know if a sample has been contaminated and therefore is unreliable. How do you know if a volcanic eruption has contaminated a sample that is supposedly 20,000 yrs old?
I suppose by this you mean the idea that a volcano erupting close to a forest can give out large quantities of C02, which can be absorbed by them and thus give a false reading. Proposed first by Woodmorappe, I believe.
Dead easy to spot. Remember trees are effectively alive only on the outside of the wood - the current growing ring. For all other growth rings, no more CO2 occurs. As a result, for carbon dating samples are taken from the rings - to date the tree's death from the outside. This can be checked by taking samples progressively further back to work back to the tree's initial seeding. If there was such an anomoly, the carbon dating results (from the inside out) would show a smooth age progression, then one (perhaps a few more) totally off the wall, followed (assuming the tree survived) by a resumption of the smooth trend.
However, I'm not convinced this is a reasonable scenario. Even the most massive CO2 surge would very swiftly be diluted in the atmosphere. The tree would have to be virtually on top of the volcano: in which case it would almost certainly be killed before it had time to take up much (if any) of the CO2. Further, even given the survival of the tree, the cloud would dissapate so fast that only a small amount of anomolous CO2 would be taken up.
So, in summary, my answer is the carbon dating would show an obvious anomoly if the event occured, but the liklihood of the event is, in any case, small.