Accuracy of carbon dating.

An interesting read, thought you guys might find it interesting.

http://www.christiananswers.net/q-aig/aig-c007.html

At first glance, it's got a big focus on C14 dating, and not much on other methods, like PbU etc., which are used for much older things. There's also a whole bunch of stuff in there that's blatantly wrong; red blood cells in dinosaur bones, for example, and why the sea should be saltier. I would take that whole lot with more than a pinch of salt...

And if I recall, they've benchmarked carbon dating pretty accurately.

Claymore said:

At first glance, it's got a big focus on C14 dating, and not much on other methods, like PbU etc., which are used for much older things. There's also a whole bunch of stuff in there that's blatantly wrong; red blood cells in dinosaur bones, for example, and why the sea should be saltier. I would take that whole lot with more than a pinch of salt...

Click to expand...

I don't really know the first thing about dating techniques.. but I've always wondered something:

How do you "calibrate" a technique to determine it's accuaracy over a very long period?
When I say very long I mean several thousand years or more.

Lets say we dig up an ancient roman artifact and use a particullar technique to date it and we get a result of +- 2000 years, and then we use the same technique to date something from the "Elizabethian" period and it gives us +-450 years, then I would suppose we could reasonably assume that the technique is marginally accurate at least for such periods seeing as we already knew or could make educated guesses about the approximate age of both items.

Now we use the technique on some old rock or something and suddenly we get a result of 100000 years.. .. seeing as we do not know the age of rock and cannot make any educated guesses about it's age and we have no historical items known to be that age how can we determine if the technique is still accurate? Seeing as there is a huge difference between 2000 years and 100000 years (let alone several million) can we really assume that the technique remains accurate over any period? Even a slight unoticable error over a short period could yield major inaccuracies over longer periods.

Just my thoughts, be they of no interest..

Accuracies are relevant to the timescale we're talking. 1000 years would be a pretty massive 10% problem if we're talking history of the last 10 years, but a 100000 years is only 0.01% if we're talking billion year timescales.

Claymore said:

Accuracies are relevant to the timescale we're talking. 1000 years would be a pretty massive 10% problem if we're talking history of the last 10 years, but a 100000 years is only 0.01% if we're talking billion year timescales.

Click to expand...

Hmm.. yes I see your point.. in a billion years an inaccuracy of even several thousand years could be negligible..

But I still see another problem, this assumes that the inaccuracy follows a linear path.. i.e., that an inaccuracy of 0.01% over 1000 years (0.1 of a year) still remains 0.01% at one million years. What if the inaccuracy were following an exponential path?

Using my earlier example:
Let's say that the "ancient Roman" and "Elizabethian" artifacts are in reality exactly 500 and 2000 years old respectively, our dating technique reveals the "Elizabethian" to be 501 years and the "ancient Roman" to be 2005 years old. Thus in reality we have an increasing error: 1 year over 500 and 5 years over 2000.. As I stated earlier we cannot really know the precise age of either of these items, we can only make educated guesses, as such we would still in this case assume the dating technique to be relatively accurate for these periods (which it would be more or less) but it would in fact be incorrect to think that it would be within the same percentage of error over a much longer period.

In my (hypotehtical) example we had an inaccuracy of 0.2% over 500 years yet by 2000 years we had an inaccuracy of 0.25%. Assuming a continuous increasing exponential we could be off by several percent by the time we reach a million years. Of course this is only if making the assumption that the error margin infact increases at a continuous exponenial.. What if it was an "exponentially increasing exponential" D) or even just a bit random.

The bottom line in my humble opinion:
I submit that we cannot really rely on any age presented which is significantly older than known human history as it must involve assumptions and is nigh impossible to prove.
- I hope I am not rubbing someone the wrong way by saying this.. - Not my intention.. just my personal opinion (which is not cast in iron).

redarrow said:

Hmm.. yes I see your point.. in a billion years an inaccuracy of even several thousand years could be negligible..

But I still see another problem, this assumes that the inaccuracy follows a linear path.. i.e., that an inaccuracy of 0.01% over 1000 years (0.1 of a year) still remains 0.01% at one million years. What if the inaccuracy were following an exponential path?

Click to expand...

You make good points. The answer to that one, I think, is that the inaccuracy is not necessarily a function of the age. Different methods have different levels of accuracy, and a lot could depend on the exact circumstances too.

Then there's the issue of correlation: you could date some rock at, for argument's sake, 500 mya. You may not be sure exactly how accurate that is. But then other samples from the same layer are independently tested, and you get 20 or 30 slightly different results. You can then plot a probability chart showing the cluster, and eliminate any results that are seriously out, and give you a tighter range.

The other thing, too, is the relevance. We obviously focus far more on accuracy when relating to human history, partly because we're humanocentric, but also because human history is directly relevant to us. So knowing historically important things happened at a particular time is more relevent to us than arbitrary happenings in human history, and much more relevant than things that happened 500 million years ago. An error of a few million years here or there makes no difference at all to us.

What young earth creationists are proposing, though, is that the earth is a matter of thousands of years old - typically 6000, though Mr TB seems to like 14000 lately. This is like saying that the distance from LA to New York is about 10 metres. The magnitude of that error is so massive that if correct it would completely undermine the underpinnings of physics, geology, chemistry, and biology completely; the very foundations of science as we know it would fall apart.

redarrow said:

Hmm.. yes I see your point.. in a billion years an inaccuracy of even several thousand years could be negligible..

In my (hypotehtical) example we had an inaccuracy of 0.2% over 500 years yet by 2000 years we had an inaccuracy of 0.25%. Assuming a continuous increasing exponential we could be off by several percent by the time we reach a million years. Of course this is only if making the assumption that the error margin infact increases at a continuous exponenial.. What if it was an "exponentially increasing exponential" D) or even just a bit random.
The bottom line in my humble opinion:
I submit that we cannot really rely on any age presented which is significantly older than known human history as it must involve assumptions and is nigh impossible to prove.
- I hope I am not rubbing someone the wrong way by saying this.. - Not my intention.. just my personal opinion (which is not cast in iron).

Click to expand...

HaHaHa foolish dating methods

TheBossMan said:

HaHaHa foolish dating methods

Click to expand...

Why don't you go start a petition to have the science subforum closed because science offends you by contradicting your religion instead of trolling for Jesus?

The Dating methods are fairly accurate over longer periods of time. To answer the question of whether we can say a specific rock or fossil is a suirtain age, if we have no reference. We actualy have a reference, the big bang, where all the minnirals were and their isotopes were created. Some of these isotopes have readioactive decay rates, which can be determined quite accurately. So in dating procedures we usualy compare the rarioactive istopes with the stable ones.
The whole process is a bit more complicated, but atleast we have a standard reference to which we compare.

Yeah look at my I'm a Troll

-> me don't care what u call me

me not getting mad anymore about what you think about me o rsay to me...
me just turned of the mad switch in my brain and broke the button, so me not givin a damn what u think..

TheBossMan said:

Yeah look at my I'm a Troll

-> me don't care what u call me

me not getting mad anymore about what you think about me o rsay to me...
me just turned of the mad switch in my brain and broke the button, so me not givin a damn what u think..

Click to expand...

Perhaps when you finish school and - hopefully - go on to tertiary education, you will be able to make a valid contribution to this section. Until that time though, would you please refrain from polluting this subforum as well?