Biophysicist writes:
On what assumptions or conditions does K-Ar dating rest, other than:
* that the rock being dated has not undergone a significant heating/cooling cycle since it was initially formed
* that the rock is impermeable to Ar, and so traps what Ar-40 is a result of radioactive decay and does not leach Ar-40 from the atmosphere
* that the rock is sufficently old for some buildup of Ar-40 (100K years or so)
Omit that last. You don't need to assume sufficient age; though you do need to pay attention to the error bars on any calculated age. More on this below.
The other two assumptions are variations of the same thing.
Another very important assumption is that the rock did not entrap any Argon at the time of its formation. This assumption is probably the most important source of error with K-Ar dating. There are various ways in which a rock might include Argon at the time of initial formation -- such as inclusion of solid crystals which have Argon already within their own crystal matrix. This kind of error has led to Argon-Argon dating becoming more common, which is a form of isochron dating that can handle such issues. But I digress.
Another assumption -- which is not a basis for any concern -- is the that radiactive decay rates are constant over time. It is worth mentioning this, since creationists are likely to bring it up. If you are likely to be explaining this to an audience including creationists, it would be wise to check up a bit about the many methods by which constancy of decay rates is tested and confirmed. The point here is that geologists don't do any additional special checks for decay rate variations in any particular rock. Decay rate variation is the domain of particle physics. Geologists leave testing of that matter to the physicists.
The basic assumptions can be given as follows:
- The rock started out with no Argon.
- Over time, no Potassium or Argon enters or leaves the rock; the atoms of the rock we measure are the same atoms present at formation from a molten state.
- Over time, Poteassium decays to Argon according to the known decay relation.
I'm a bit confused because Geochron labs has their minimum date on the order of 0.5M years, while some webpages I'm finding at UCSB talk about K-Ar dating possibly being useful for dating rocks as young as 20K years.
It depends on the accuracy of instruments. For example, some labs may have some residual atoms of Argon lying around from the last sample measured, and this will contribute to small errors in the next.
If the amounts of Argon in the next measurement are very small, then the final Argon count will be sigificantly due to foreign Argon. Some labs charge more for young samples, because they need to prepare the measuring equipment much more carefully.
But in any case, you should get error bars for known sources of measurement error, like Argon contamination in measurement. For a very young sample, the error bars will exceed the age... which means it is not actually useful for finding the age of the rock.
You can still do the test, and this is sufficient to distinguish a young rock from an old one. But you can't get a meaningful age. You can only say that it is something younger than some lower bound.
If you have a messy rock which is likely to have included Argon, then dates you get back may be meaningless. Geologists are well aware of this potential for error, and have various ways to judge the reliability of a date. Some kinds of rock are better than others -- and this has been tested. In general, a confirming date by some independent means is very handy.
My quick on-line check suggests that
Geochron labs does not currently do K-Ar dating. I know they have been used by creationists in the past to measure some of their samples, and they have done K-Ar in the past. The qualifications you list above were considered here in
Message 77.
I've skimmed some reports by creation "scientists" such as Steve Austin describing a mockery of this technique by using it to derive dates of up to a million years for historic lava flows. I use the words "up to a million" advisedly, because all of the supposedly erroneous dates fall nicely at the bottom end of the spectrum K-Ar is meant to test.
I also saw a defense of those techniques by some guy at AIG, but I didn't like his assertion that the error of the measurements being 60K years and the mean being ~300K years strongly indicated that it was the test itself, not the experiment, that was the problem. Is there some natural distribution to the amount of Argon present in rocks, such that we can measure the amount actually there to a finer precision than the natural variation? If so, how does this distribution arise? If not, why do we not meaure zero Ar-40 in rocks that were just coughed out of a volcano?
A lot of the errors are due to included Argon. Austin is actually someone who has real experience in this area. This doesn't make him any more sensible (just the reverse, in fact) but it does mean we need to be careful to check what the source of errors might be.
You can check out some critiques of some of Austin's work at talkorigins; but if you have a specific claim from Austin for which you don't know some careful analysis from someone well informed in geology, then you need to be a bit careful about identifying the problems. It would be a great topic here at EvC.
We are at present suffering an infestation of people who want to overthrow conventional science but have no idea at all of the matters they purport to refute. Many of the posts look like Finnegan's Wake on drugs. We are all starving for a creationist argument with some meat on it.
Cheers -- Sylas
[This message has been edited by Sylas, 03-30-2004]
™ Version 4.2