Hacker News Reader

Comment by apunic | Hacker News Reader

I always wonder how scientists do know that something was 25, 540 million or 2.560 billion years ago.

Single crystal uranium-lead dating is one of the best all-around methods and gets used a lot for vanilla geological applications of Earth rocks; any rock that contains trace amounts of the mineral "zircon" can be dated with this technique. Zircon is special because it tends to incorporate a lot of uranium when it crystallizes while excluding lead. As time passes after the zircon crystal cools, uranium decays into radiogenic lead, and by measuring the relative abundance of these two things in the crystal we can estimate the time since that zircon crystallized (with no lead inside) and get an "age" that tells us how long it has been since that zircon crystal was formed. Zircon is common in many types of igneous rocks, which are rocks that began in a molten state. The zircon crystals in an igneous rock are resistant to weathering (compared to feldspars, another common igneous constituent), so sometimes they end up as grains in sedimentary rocks and we can use a variation on the technique called "detrital zircon geochronology" to estimate the age of sedimentary deposits.

Other systems (like samarium-neodymium) are most often used when trying to date things older than the Earth, e.g. meteorites.

http://en.wikipedia.org/wiki/Radiometric_dating

Carbon Dating

Wrong, carbon dating has a limit of 62 millenia. Maybe you meant radiometric dating?

Of course I did.

With out knowing, I some how got involved with carbon dating. I must have, I married a diamond.....

Two reactions to that, either: ahhhhh sweeeeeeet, or I feel sick, pass me a bucket. So, up vote for sweet, or down vote for bucket. :)

(Normally I like a Sunday rant. I feel different today)

Is this joke yours? It's great!

Carbon dating is actually only effective up to about 50,000 years, so its really only ideal for tracking human history after we started migrating out of Africa. This is because the unstable carbon-14 isotope only has a half-life of about 5700 years.

There are many other forms of radiometric dating, though. Uranium-lead dating has proven relatively accurate for periods of time between 1 million to 4.5 billion years ago.

Note: I'm not a geologist. I remembered that carbon was only effective for a few thousand years from a geology course I had in college, but all of this information is actually taken from articles on wikipedia: http://en.wikipedia.org/wiki/Radiometric_dating

I'm curious how it could prove accurate for that time frame? How could something like that be verified or proven?

Due to the way radiometric dating works, you don't need to actually verify in the way that you are using the word. However, there are at least two ways to verify the date:

1) Verify the half-life value. Radiometric dating works on the same principle: look at the ratio of pre-substance to fissioned substance. If you have a good half-life value, you can determine the age of the specimen using those values.

2) Use two different methods for getting the same age. If you use multiple, independent methods of determining the age of something, and they all come out the same, you have greatly lessened the probability that you're wrong or missed something.

Of course there are ways that these can be strengthened even more using practices like double blinding the measurements.

It's all theoretical, in the same way that our understanding of gravity or radiation itself is theoretical.

By that metric, our understanding of anything is theoretical.

Yep.

It's more like a , you might get killed by a car if you step into the street kind of theoretical. Or the drip of a water clock will fill the container in 8 hours kind of theoretical. Or a don't plant too early or your people will starve kind of theoretical.

A theory in science is a model backed by the huge majority of observational evidence and the collective experience of the model's community. There are no other equivalent human institutions for reliability. None.

So saying "it's just a theory" in the case of uranium-dating is a fairly unreasoned attitude.

I'm not saying it dismissively. "It's a theory" is not a negative thing to say. It does not mean that I doubt it in any way.

The best way I know to "check" radiometric ages is to date a rock using more than one isotopic system (e.g. K-Ar and U-Pb) and compare the ages that you get. If you get the same age from isotope systems with different decay rates and systematics, that is strong evidence that your technique is working.

People have done this many times, and most of the error comes from different levels of uncertainty in the decay constants from one system to another. Many of the decay constants have not been updated for decades (e.g. Steiger and Jager 1977) and are suspected to be a few percent off, but the people who are qualified to make such measurements (physicists + chemists) aren't the people who want to use them in applications (geochronologists and geochemists). So the incentives aren't aligned and I doubt anyone will ever fix them.

Note: Was a geology/geophysics PhD student until recently.

Reply on news.ycombinator.com