Age of Stars: Calculating & Assessing Helium

[adkinje]

Is it correct that the age of a star, such as the sun, is calculated by measuring the amount of helium made from fusion? If so, how do we know if that the star didn't form from a cloud that already had helium in it?

 

[mathman]

Right after the big bang, H1 made up about 75% of ordinary matter and He4 about 25%. The remaining primordial species (H2, He3, Li6, Li7) combined were less than 1%. To answer your question, what is looked for is an increase in He4 / H1 ratio from 1/3.

 

[twofish-quant]

Not really. What you do is to run computer simulations to see how the star changes over time. It's rather difficult to use this to measure the age of a single star, and this is most useful when you have a lot of different stars that are at in a cluster.

The amount of helium for a star doesn't change that much because all of the helium that the sun creates is still buried in the center of the star. We can tell that the sun is a third or fourth generation star because its surface contains heavy elements that the sun did not create.

The main limits on the age of the sun come from radioactive measurements of meteorites.

 

[Wallace]

Measuring ages of stars is quite a tricky thing to do. As TQ said, you can generally do much better by looking at the age of a cluster of stars rather than an individual. Even in that case the errors are significant.

 

[twofish-quant]

Also just to be clear. When Wallace says "significant errors" what he is saying is that by observing a star we can tell that the age is "several billion years" but we can't tell exactly how many billions without a lot of detail.

Some pretty simple math (how much fuel does the sun have, how fast is it burning it) will give you that the sun is several billion years old and not tens of billions or hundreds of million years ago. This matches meteorite measurements that say that the solar system is 4.5 billion years old. If you want details, then you have to start taking into account a lot of complicate effects (like how much mass is the sun shedding, how to take into account mixing, etc. etc.)

If the meteorites said that the Earth was several billion years old, but it seemed that the sun was several tens of millions, then we'd have a problem. This actually happened in the late 19th century, when the geologists insisted that the Earth was several billion years old, but the astronomers said that the sun couldn't be more than a few tens of millions of years old, because that's how much the sun energy the sun could emit by gravitational contraction.

It turned out that the geologists were right, because what astronomers didn't know was nuclear fusion, and it was in the 1920's that a new source of energy solved the problem.

 

[bombadil]

I've always had the impression that one can arrive at ~4.5 Gyr for the age of the sun via simulations/modeling. I thought dating meteorites just pinned down the age of the solar system/sun with greater accuracy. Is this correct?

 

[twofish-quant]

You can get an age of "several billion years" by simulation. The basic problem is that while the sun is on the main sequence, the star doesn't change very much and so its hard to tell if a star like the sun is 3 billion or 6 billion just from the model since stars like the sun look about the same if they are 3 billion or 6 billion years old. (They start looking *really* different at 1 billion or 10 billion). There are things in the model that you can change to get a young sun or an old sun.

In practice it works the other way, since we know from meteorites that the sun is 4.52+/- 0.04 Gyr, we change the models so that we get that as the exact of the sun, and then use those models to look at other stars.

It's very hard to tell the age of a single 1 solar mass star. It's much easier to tell the age of a star cluster. The thing about stars like the sun is that it's hard to tell if they are 3 or 6 billion years. If you have a 5 solar mass star, they only live for 100 million years before they change a lot, so if you have a star cluster, you see what the high mass stars look like, and that will let you pin the age of the cluster.

 

[twofish-quant]

Also this also means that when the geologists say ooppss... We thought that the meteorites said that the solar system was 4.7, but not we think that it's 4.52, that people go in and tweak their models to match the new age.

The sun is really important since we have a huge amount of information about it so people need to change their models so that when you put in the sun, you get the right numbers. One thing that gives us huge amounts of information is astroseismomology which sees how sound waves goes through the sun and can get us temperature and density profiles.

http://www.ap.stmarys.ca/~guenther/Level01/solar/what_is_ssm.html

 

[qraal]

There's a minor furore occurring in solar modelling due to changes to the estimates of the Sun's core metallicity, specifically oxygen levels (as I remember), thanks to recent updates in the observed elements in the solar wind. They're wrestling with the numbers trying to get a nice harmonious match between the models and the helioseismology data, like John Bahcall managed with the old metallicity levels.

Also this also means that when the geologists say ooppss... We thought that the meteorites said that the solar system was 4.7, but not we think that it's 4.52, that people go in and tweak their models to match the new age.

The sun is really important since we have a huge amount of information about it so people need to change their models so that when you put in the sun, you get the right numbers. One thing that gives us huge amounts of information is astroseismomology which sees how sound waves goes through the sun and can get us temperature and density profiles.

http://www.ap.stmarys.ca/~guenther/Level01/solar/what_is_ssm.html

 

[bombadil]

I haven't had much exposure to astroseismology, any recommended introduction/link for the curious with a background in physics such as myself?

 

[twofish-quant]

Cool... One thing that happens a lot in science involves increasing the accuracy of models. There is something of a consensus that the solar system is between 4 and 5 billion years old. But once you have that, then you want to decide if it's 4.1 or 4.8. Once you have 4.5, then you want to know if it is 4.51 and 4.58?

 

[Chronos]

Mass and composition determines stellar ages. The age estimate becomes pretty good once you nail those properties down.

 

[Wallace]

Mass and composition determines stellar ages. The age estimate becomes pretty good once you nail those properties down.

Knowing the mass, luminosity, metallicity of a star doesn't tell you the age. It does tell you which simulation result to go to in order to get the predicted age from models, but as explained in this thread, those models have reasonably large uncertainties (although the level of uncertainty is different for different types of stars).

So yes, age estimate improve once you know those stellar properties (which is what you would expect, since before you know them, what do you actually know about a given star??) but whether or not you think the estimate is them 'pretty good' depends on your tolerance for uncertainty!

As twofish-quant explained, the only age estimate that is really trusted to any great degree is looking at the location of the turn-off in a colour-magnitude diagram in a cluster. This still gives you an absolute age with a reasonably high uncertainty, but if you can get a decent distance estimate to the cluster (in order to map the observed luminosities to absolute luminosites) then you can get quite an accurate relative measure of the ages of different clusters.

 

[AdkinsJr]

It seems that astrophysicists have accurately measured the age according to the abstract of this paper.

http://www.aanda.org/index.php?option=article&access=doi&doi=10.1051/0004-6361:20020749

It's pretty extraordinary that they got 4.57 +\-n 0.11 Gyr.

 

[twofish-quant]

It seems that astrophysicists have accurately measured the age according to the abstract of this paper.

It's actually not that impressive. Looking over the paper, I can count at least four or five fudge factors which you can tweak in order to get the right age. Right now people are still using the known age of from meteorites to calibrate the models, rather than going from the models to get right age.

Looking at the paper, if I didn't tell the author what the age of the meteorites were, there is enough wiggle room to probably move the age of the sun about a half a billion years or so.

The important thing about the fudge factors is that a lot of them don't change from star to star. Once we fix a good number for the fusion reaction rate s_pp , then we have to use the same number for models of other stars.

 

[AdkinsJr]

Not really. What you do is to run computer simulations to see how the star changes over time. It's rather difficult to use this to measure the age of a single star, and this is most useful when you have a lot of different stars that are at in a cluster.

The amount of helium for a star doesn't change that much because all of the helium that the sun creates is still buried in the center of the star. We can tell that the sun is a third or fourth generation star because its surface contains heavy elements that the sun did not create.

The main limits on the age of the sun come from radioactive measurements of meteorites.

Is there any way to know if the heavier elements were present when the sun formed? Isn't it possilbe that they could have been transported to the sun via asteroids, comets, etc. after it had already formed?

 

[twofish-quant]

Is there any way to know if the heavier elements were present when the sun formed? Isn't it possilbe that they could have been transported to the sun via asteroids, comets, etc. after it had already formed?

Figuring out what happened is like putting together a jigsaw puzzle. So a lot of the reasons that you come up to explain something depend on other pieces of the puzzle being more or less in the right place. The problem with having the sun have heavy elements after formation is that 1) you don't have enough mass in the asteroids and comets to change the composition of the sun much and 2) you have to explain where the asteroids and comets got their heavier elements from.

 

[Chronos]

Nucleosynthesis is a fairly accurate science these days. You don't need astronomical observations to make that assertion, particle accelerators have already made the case. We know the temperatures required to synthesize heavy elements and know the kind of stars necessary to achieve these temperatures. I fail to see the issue on this point. Transport models fail in the face of the vast distances required to get them 'here and there' and still make sense in the face of evidence already collected.