Jeremy Pallant

The Space section of CNN.COM has an interesting item in it today, in which the discoverers of a quasar 13.5 billion light years away reveal that it contains much more iron that it should, something which has interesting implications for the age of the universe.

The full article is <a href="http://www.cnn.com/2002/TECH/space/07/10/universe.age/index.html" target="_blank">here</a>.

fando

Even if the universe were older, it still wouldn't explain why the quasar shows Fe concentrations three times that of our solar system. The starstuff our solar system is made from got recycled at least once in the history of the Galaxy, so we should have higher Fe concentrations than that quasar no matter what. Personally, I'd call for a second opinion on those measurements and then seriously consider whether there can be exotic factories of metals in the early universe. Such objects would affect a horde of metal concentration dependant assumptions in Astronomy and Astrophysics.

Boro Nut

I heard that too. Weren't they the Tampa Bay Rowdies before Pele signed?

Boro Nut

Answerer

It seems that it is either the red shift methods is wrong or there is still answers needed to be found. Somehow, I don't wish red-shift method is be falsified as this will give our current knowledge of physics lots of problems.

Tim Thompson

Don't jump on the PR bandwagon too fast. Press releases almost never tell the right story, and this case is no different. The actual research paper, already published, is on the web, and far more informative than the PR blurb.

<a href="http://xxx.lanl.gov/abs/astro-ph/0207005" target="_blank">Discovery of an ionized Fe-K edge in the z=3.91 Broad Absorption Line Quasar APM 08279+5255 with XMM-Newton</a>
G. Hasinger, N. Schartel & S. Komossa
Astrophysical Journal 573: L77-L80, 2002 July 10

The Fe abundance is pinned down by locating an "edge" like feature that it would create in the X-ray spectrum of the quasar. Under normal circumstances (no Fe), the spectrum would be well fit by a power law that depends on absorption in our own galaxy & the quasar host galaxy. But the power law fit in this case is "statistically unacceptable", and the residual plot shows the edge. If you look at the spectrum by eyeball, the edge is by no means "obvious". So right off the bat, you understand that this is not a "strong" detection. furthermore, it is the only Fe "edge" yet detected, so it is only one example of one quasar, a single datum. The result will not be very convincing, as far as I am concerned, until we see a few more, and until we can get away from the "excess residuals" argument.

Meanwhile, the result is used to bootstrap an iron to oxygen ratio (Fe/O), which is found to be 3.3±0.9 of the solar ratio, in the third & longest <a href="http://sci.esa.int/home/xmm-newton/index.cfm" target="_blank">XMM-Newton</a> data set (a <a href="http://chandra.harvard.edu/" target="_blank">Chandra</a> data set shows no feature, and a shorter XMM-Newton data set is inconclusive as well).

In the <a href="http://sci.esa.int/content/news/index.cfm?aid=1&cid=1&oid=30255" target="_blank">ESA PR blurb</a>, one of the scientist says "One distinct possibility to explain these observations is that, at the redshift we are looking at, the Universe is older than we think." But, in the paper they say "Furthermore, at a redshift of z~4 the age of the universe is a little less than 1 Gyr (for q0 = 0.5), which compares to the timescale of ~1 Gyr necessary to enrich Fe/O up to the solar value. Given that we find strong indications of a supersolar Fe/O abundance, we are beginning to constrain cosmological models, favoring those that predict larger galaxy ages at a given z."

Older universe or older galaxies? They don't necessarily mean the same thing. Aside from the age of the universe being a bit of a free parameter (it depends on the "deceleration parameter", q0, amongst other things), so is the rate of star formation in the very early universe, so is galaxy formation in the very early universe, and so is the chemical evolution of galaxies (due to star formation), in the not so early universe. So there is a lot of padding available before we start sacrificing the age of the universe to the God of excess residuals in a spectral fit to a single quasar.

But even if the result & conclusions stand, it hardly affects the "redshift method". All it does is reset the redshift - distance relationship, such that the universe is older for a given redshift, such that objects are more distant from us for a given redshift. That's a detail, and not a fundamental problem. Even if true, it's a minor perturbation on Big Bang cosmology, nowhere near as significant as the PR blurb, or news report, would lead one to believe.

fando: The starstuff our solar system is made from got recycled at least once in the history of the Galaxy, so we should have higher Fe concentrations than that quasar no matter what.

So it might seem, but it ain't necessarily so. In relatively small stellar systems like our own, most of the Fe never makes it out of the core, and is destroyed in a <a href="http://astrosun.tn.cornell.edu/courses/astro201/sn.htm" target="_blank">supernova explosion</a>. However, in a supermassive system with deep convection you can skew the Fe/O ratio by dredging up Fe from the core before it goes boom. It might well be that you expect a higher Fe/O ratio for a quasar like system, for just that reason. So until somebody models that kind of process in more detail, I should think that's not a done deal either.

fando

Oh, I was thinking of the ambient metallicity. My comments aren't really relevant anyway, I didn't look into the matter deeply as you did. If the quasar is dredging up metals through convection, then that's something new to me and would fit my suspicion that there's a means of pumping up the apparent metallicity of quasars.

And I agree that we need more data. The media never seems to differentiate between exciting possibilities and time tested theories. Every article is about a new development that challenges the establishment and revolutionizes our views.

beausoleil

I'd like to understand this, but I'm afraid I don't. You seem to be comparing a quasar - a black hole at the centre of a galaxy (essentially a galaxy core) with our stellar system (our star?) - but our star isn't going to go supernova - it isn't big enough. In fact, I don't think it is even going to get to the iron synthesizing stage.

In fact, isn't much of the galaxy's iron (predicted to be) made in type 1a supernovae? Might that not explain a high iron content in a galactic core, since 1as are in a sense a collision process. I'm speculating that as white dwarfs encounter the quasar black hole's accretion disk they absorb enough local material to push them over the explosion mass. I think we expect about a solar mass of iron per 1a supernova (from memory).

On the other hand, I think somebody - was it Wasserburg? - suggested the first generation stars were supermassive and processed large amounts of material to the iron peak in an Ap J paper a while back. Maybe that will be resurrected if the result stands.

Tim Thompson

beausoleil: ... but our star isn't going to go supernova - it isn't big enough. In fact, I don't think it is even going to get to the iron synthesizing stage.

Our sun is far too small to make any iron, or a supernova. It will become a white dwarf of about 0.85 solar masses, probably a He white dwarf, maybe a C/O white dwarf.

What I was thinking is that in massive stars that wind up as type II SN, you can dredge up iron from the core before the SN. If you do make iron in an accretion disk in a QSO system, the same thing might happen, since the disk is turbulent.

beausoleil: In fact, isn't much of the galaxy's iron (predicted to be) made in type 1a supernovae?

I had never even thought of that, but indeed that seems to be the case. I had written my previous note under the impression that all iron came from the core of type II SN. However, on seeing your comment I looked it up. The type Ia supernova generates copious amounts of 56Ni, which is unstable and decays to 56Co which decays to 56Fe, the primary isotope of iron (91.754%). SN 1987a created about 0.07 solar masses of 56Ni, all of which must have decayed to 56Fe (Advanced Stellar Astrophysics, William K. Rose, Cambridge university Press, 1998, p.229 & p.348). So it seems a safe bet that most of the iron around came from type Ia SN.

This makes the issues that I rasied before even more important than I thought they were! The rate of star formation will certainly affect the rate of type Ia SN, which in turn makes the abundance of Fe even more sensitive to the star formation rate.

As far as I know, everybody in the game thinks that the first generation of stars was supermassive, probably 100 - 300 solar masses. Evidently there aren't any stable single star solutions above about 1000 solar masses in any case. But i don't know the rate at which they might synthesize iron. No doubt that's another of those free parameter mystery iron factories we have to worry about.