# Mature stars in ancient galaxies

Gold Member

## Main Question or Discussion Point

http://arxiv.org/PS_cache/astro-ph/pdf/0502/0502385.pdf [Broken]

Hence the galaxies SBM03#1 are already, to some ex-
tent, established systems. This is a significant finding since,
at z ≈ 6, the Universe is less than 1Gyr old.
No kidding. Would anybody here like to hazard a guess at where this "reddening dust" lies on the path between the z~6 galaxies and our vantage point? If the "reddening dust" is very ancient and distant, I would be interested in hearing how it could have been formed in the low-metallicity environment predicted by the BB theory. If it is more local, is there a model explaining the preferential reddening of distant galaxies?

Light should follow the same rules in distant environs as it does here. If not, the universe would look pretty strange. When we observe trends in the behavior of light, we should first explore the assumptions regarding the propagation of light in a "vacuum" before positing the existence of intervening materials, like "reddening dust", "iron fibers" (another weird but common one!) etc.

When light enters a denser propagating medium, it slows down, and when it encounters a less-dense medium, it speeds up. When it encounters a density transition that is not perpendicular to its direction of travel, it is bent. This is classical optics. Space is not empty, boys and girls. Space is a transmissive medium. That medium may not have boundaries as well-defined as those between the air/water surface of a swimming pool, but the properties of the transmissive media must be taken into account in any classical optical application, and we have not yet properly considered the role of the quantum vacuum as the transmissive medium (ether) through which EM propogates. If we are not willing to take this conceptual step, I believe that physics is going to be stalled for a very long time.

rant mode OFF

Last edited by a moderator:

Related Astronomy and Astrophysics News on Phys.org
what are you on about? read section 3.5, they conclude that the models are best fit with no reddening due to dust.

Chronos
Gold Member
I vote with matt o. Dust models have long since been ruled out in the grand scheme of redshift corrections.

Gold Member
Chronos said:
I vote with matt o. Dust models have long since been ruled out in the grand scheme of redshift corrections.
Apparently not with these folks, since they made several critical comparisons with dust-reddening models in the course of their research. I'm happy that they managed to get good fits in their model without dust reddening, but am perplexed about how much significance they attached to it, if the idea is truly discredited.

I'm sorry, but I think you really need to re-read the paper. The Calzetti reddening law is empirical and is proportional to the wavelenght of the light. They merely investigate the reddening to show the redness is caused by an older population of stars, not a dust reddened starburst. The reason they do this is because starbursts are known to be dusty, with intrinsic reddening increasing with star formation rate. They are trying to show that these galaxies had an older population of stars in place at this redshift, hence it is a relatively mature galaxy (for this epoch). They conclude that there must have been starbursts at an earlier epoch which may have helped reionise the early universe.

Last edited:
Gold Member
I understand.

They also constrain the upper age of the mature star populations by saying that our theory says that the universe is 13.7GY in age and at z~6 the galaxies are less than a billion years old, thus the apparently mature stars must be under a billion years old. There are some models (Garth's included) in which the universe is considerably older than that, so it might be a good idea to constrain theory with observation, and not the other way around.

SpaceTiger
Staff Emeritus
Gold Member
turbo-1 said:
I understand.

They also constrain the upper age of the mature star populations by saying that our theory says that the universe is 13.7GY in age and at z~6 the galaxies are less than a billion years old, thus the apparently mature stars must be under a billion years old.
No, that's incorrect as well. They determine a range of possible ages for the stellar population (250-650 Myr) by fitting stellar models to the spectrum. These are not dependent upon cosmology.

What is dependent upon cosmology is the formation redshift that these ages correspond to (z ~ 7.5 - 13.5). If early stars are responsible for reionization, then this is entirely consistent with the WMAP results, which imply that stars were forming as early as z ~ 17.

turbo-1 said:
I understand.

They also constrain the upper age of the mature star populations by saying that our theory says that the universe is 13.7GY in age and at z~6 the galaxies are less than a billion years old, thus the apparently mature stars must be under a billion years old. There are some models (Garth's included) in which the universe is considerably older than that, so it might be a good idea to constrain theory with observation, and not the other way around.
Yes, they do. As Spacetiger has poined out, other observations point to a 13.7 Gyr old Universe. I don't see how a star-formation model could be used to accurately constrain the age of the Universe and overthrow the current models. As you can see from the paper, there are enough degeneracies in these models as it is. We don't actually know much about massive star formation, so using it to constrain cosmological parameters seems a bit far fetched!

Gold Member
OK, what does this mean?

paper said:
Figure 10. The allowed range of masses for several exponentially-
decaying SFR models for SBM03#1, with decay times ranging
from  = 10−1000Myr, as well as an instantaneous burst model
and a constant SFR model. Those with stellar ages > 1Gyr (right
shaded region) are ruled out (the Universe is younger than this at
z ≈ 6).

What I am trying to say is that these models are ruled out based on other observational data. They are likely to be caused by some other degeneracy within the model, they should not be taken to mean the universe is older than 13.7Gyrs. You cannot make a SFR model with every parameter left to vary, there are too many degeneracies, hence you set the known parameters and constrain the unknown.

SpaceTiger
Staff Emeritus
Gold Member
turbo-1 said:
OK, what does this mean?
Alright, you're correct, I overlooked that. What matt.o is saying is also correct, however. Fitting z = 6 spectra is extremely difficult without making some assumptions. :yuck:

Gold Member
SpaceTiger said:
Alright, you're correct, I overlooked that. What matt.o is saying is also correct, however. Fitting z = 6 spectra is extremely difficult without making some assumptions. :yuck:
Understood. If the LBT and the Webb allow observations of galaxies at z~8-10 and mature stellar populations are observed at those redshifts, what then? Do we simply ramp back the rapid star-formation epoch to z~22 and press on? That's the current trend, and it's disturbing.

We have already reached a point where observations should have made us reconsider our assumptions. We know that objects at z~6 exhibit solar and super-solar metallicities with no observed metallicity evolution with redshift. This makes perfect sense in a steady-state universe, but it is harder to reconcile with a Big Bang model in which metallicity must evolve - even in Garth's model, which is freely coasting and provides several billion more years of breathing room at the front end.

We do need to make assumptions, but I'm concerned that our assumptions are in need of some sober reconsideration.

Your first paragraph is just irrelevant, as is the third. As for the middle paragraph, you seem to have failed to notice that the best fit models were the lower metallicity ones. Anyhow, I don't see the problem with high metallicities at z~6. We are obviously looking at objects undergoing extreme bursts of star formation and other activity. These bursts can last a few million years and heavily pollute a galaxy with metals very quickly. The extremely high star formation rates in these bursts could easily produce a population of older stars whilst the larger stars go supernova and pollute the ISM with metals.

Surveys out to z~6 are hardly complete and are extremely biased to the brightest of galaxies, so I don't see how you can make broad assumptions about metallicities and star formation of the universe at these redshifts. There is definitely not enough evidence to overthrow current cosmological models in favour of less likely ones.

Last edited:
Garth
Gold Member
Yet matt.o and SpaceTiger, things are not that simple with the standard model; putting in some numbers:-

Using Ned Wrights calculator in the standard LCDM model, where tz=x to be the age of an object now observed at a red shift x, we have for time after BB:

For "re-combination" - the surface of last scattering of the CMB, z = 1089,
tz=1089 = 378,000 yrs.

for the onset of metallicity, i.e. Pop III stars, z = 20
tz=20 = 182 Myrs.
or SpaceTiger’s suggested figure of z = 17?
tz=17 = 229 Myrs.

for quasar 'ignition' z = 8 ??
tz=8 = 652 Myrs.

From that paper Spitzer Imaging of i'-drop Galaxies: Old Stars at z~6, for galaxy formation in the mass range 1.3-3.8x10^10Msolar z = 7.5 – 13.5
tz=13.5 = 318 Myrs. - tz=7.5 = 710 Myrs.

for 'modern' metallicity in Quasar SDSS J1030+0524 z = 6.28
tz=6.28 = 896 Myrs.

The real problem with star formation at high red shift is to explain how under gravitational collapse the
$$\frac{\delta\rho}{\rho} = 10^{-5}$$
at last scattering became
$$\frac{\delta\rho}{\rho} = 10^{26}$$ or so,
in the early stars in only ~200 Myrs?

At z = 13.5 the average cosmological density is around 10-26 gm/cc and the mass of the observed galaxies ~ 1044 gms, so that mass would have had to be gravitationally scoured from a volume 1070 ccs, i.e. a radius ~ 135,000 lgt. yrs, in about 300 Myrs. i.e. at an average infall velocity of 0.05%c!

The earlier mature galaxies are discovered the more difficult it becomes to explain their formation in the time available.

Garth

Last edited:
Garth
Gold Member
Garth said:
At z = 13.5 the average cosmological density is around 10-26 gm/cc and the mass of the observed galaxies ~ 1044 gms, so that mass would have had to be gravitationally scoured from a volume 1070 ccs, i.e. a radius ~ 135,000 lgt. yrs, in about 300 Myrs. i.e. at an average infall velocity of 0.05%c!
This may not be impossible, let us look further at the numbers. Without a detailed model of early galaxy formation we can only do an OOM calculation, but that itself may be illuminating.

First approximation: consider a steady in-fall acceleration. Note the real formation began at a very low acceleration and ended with a much higher one than in this approximation.

What constant acceleration would be required to form the observed galaxy in the mass range 1.3-3.8x1010Msolar in about 300 Myrs?

The infall distance was calculated to be 1.5x1023 cms and the time available is 3x108x3x107 secs i.e. ~ 1016 secs.

Now $$s=\frac12at^2$$, so $$a=\frac{2s}{t^2}$$, and putting in our numbers we get:
$$a=\frac{3.10^{23}}{10^{32}}=3.10^{-9} cm.sec^{-2}$$,

Reaching out to a range of 1.5x1023 cms this would require a dimensionless Newtonian potential of

$$\frac{GM}{rc^2}.\frac{c^2}{r}=\frac{GM}{r^2}=3.10^{-9}$$ so

$$\frac{GM}{rc^2}=3.10^{-9}.\frac{r}{c^2}=\frac{3.10^{-9}.1.5.10^{23}}{10^{21}}=5.10^{-7}$$
This is actually equal to the dimensionless Newtonian gravitational potential of our own galaxy's field!!

However the early galaxy is estimated to have a mass an order of magnitude smaller than our own, therefore a somewhat longer time than 300 Myrs. would appear to be required to allow it to form.

Just as a side note, and not wanting to bore anybody, the SCC/freely coasting time scales are about three times longer in this early epoch than the standard model.

Garth

SpaceTiger
Staff Emeritus
Gold Member
To the best of my knowledge, the standard model has no trouble producing galaxies at z ~ 6. If you have a reference that suggests otherwise, please provide it.

However the early galaxy is estimated to have a mass an order of magnitude smaller than our own, therefore a somewhat longer time than 300 Myrs. would appear to be required to allow it to form.
Cold dark matter scenarios are "bottom-up", meaning that the small things form before the big ones. I suggest you read up on the Press-Schechter model of hierarchical growth. The original paper is here. Much has changed since then, but that paper outlines the basic approach. For updates, feel free to peruse the enormous list of citations.

Chronos
Gold Member
I'm guessing Garth is thinking about Jeans mass issues... an intersting topic. Any ideas about zero metallicity models? I am very interested in that subject. Dark matter adds to the fun. Curious, ST, I lean toward bottom up models, but I kind of like enormous black holes confusing the issue too. Is there a good way to constrain those models? Bear in mind I'm still a bit alarmed by 512 kev photons from the galactic core.

SpaceTiger
Staff Emeritus
Gold Member
Chronos said:
Curious, ST, I lean toward bottom up models, but I kind of like enormous black holes confusing the issue too.
A good paper on early black hole growth can be found here. The high-redshift quasars are not inconsistent with current theory, but they are on the extreme end of it. As has been discussed previously, this could just be a selection bias.

Is there a good way to constrain those models? Bear in mind I'm still a bit alarmed by 512 kev photons from the galactic core.
I'm not quite sure what you mean. Are you referring to the radiation hypothesized to be due to dark matter annihilation?

Garth
Gold Member
SpaceTiger said:
To the best of my knowledge, the standard model has no trouble producing galaxies at z ~ 6. If you have a reference that suggests otherwise, please provide it.
Agreed for galaxies at z = 6, however I was simply supporting Turbo-1 and my comment that
The earlier mature galaxies are discovered the more difficult it becomes to explain their formation in the time available.
The reference is that referred to by Turbo-1 above: Spitzer Imaging of i'-drop Galaxies: Old Stars at z~6 which states
Exploring a range of population synthesis models indicates that the average stellar age is > 100Myr; our best-fit models suggest preferred ages of 250−650Myr for an exponentially-declining star formation rate (of decay time _ ≈ 70 − 500Myr) or a two-component model (with an ongoing starburst responsible for 0.5 − 5% of the total stellar mass). This implies formation epochs of zf ≈ 7.5 − 13.5 for the galaxies SBM03#1.
There is a problem in explaining the formation of 1010 Msolar galaxies at
z =13.5!

Chronos you are right in that I was thinking about the Jean’s mass, but primarily I was concerned about the time taken for such a mass to condense. Matt.o I think you must be mistaken, I only used a galactic mass 1010 Msolar in order to get a crude first approximation as to the lower limit of the time required. This can be illustrated more clearly using the free-fall time scale of a collapsing sphere that has satisfied the Jean’s criterion under homologous collapse.

$$t_{ff}=\sqrt{\frac{3\pi}{32G\rho}}$$

for $$\rho=10^{-26}$$ (cgs units) we have tff = 0.67 Gyr.

So there is not enough time for such a body to form, even at their lower end of the z scale z=7.5 (0.71 Gyr)!

Garth

Last edited:
Garth
Gold Member
We can be a little more accurate than that first approximation, although this is still a naive determination of a lower limit of the time for formation of objects in the early universe.

Using the standard model parameters the average cosmological density at z = 13.5 is

$$\rho=7.9.10^{-27} gm/cc$$ this gives a tff=0.75 Gyr..

At z = 7.5 the average cosmological density is $$\rho=1.6.10^{-27} gm/cc$$ this gives a tff=1.6 Gyr. making the situation worse!

Which shows the problem of forming anything, you have to start early.
Edit: These densities, of course are over-densities and in these naive calculations it is assumed that Jean's collapse is initiated when $$\frac{\delta\rho}{\rho} = 0.5$$.

Just for information in the SCC theory if you subsitute the density at z in tff and equate it with the universe's age at z you get a naive estimation of objects forming under Jean's mass homogolous collapse at z = 10.1. Edit: Furthermore, if we assume Jean’s collapse begins shortly after last scattering at z = 1000 when T = 3000K this model produces a Jeans Mass of 4 x 1010 MSolar that ends collapse at around z = 10 – exactly in the middle of that paper’s ball park!

Garth

Last edited:
SpaceTiger
Staff Emeritus
Gold Member
Garth said:
The reference is that referred to by Turbo-1 above:
No, the reference above is simply about observations. I'm asking for a reference suggesting that there's a problem with CDM predicting these galaxies. Your calculations are not sufficient, as they're completely incorrect.

There is a problem in explaining the formation of 1010 Msolar galaxies at
z =13.5!
The key thing to notice here is that they're suggesting the stars formed at that epoch, not necessarily the galaxy. Remember, in a bottom-up scenario, small things form first.

Chronos you are right in that I was thinking about the Jean’s mass,
The medium is not pressurized, so the Jeans mass is meaningless.

This can be illustrated more clearly using the free-fall time scale of a collapsing sphere that has satisfied the Jean’s criterion under homologous collapse.

$$t_{ff}=\sqrt{\frac{3\pi}{32G\rho}}$$

for $$\rho=10^{-26}$$ (cgs units) we have tff = 0.67 Gyr.
This is incorrect as well because it ignores the accretion of matter onto the halo. This issue was directly addressed by Jim Gunn back in 1972. See here.

So there is not enough time for such a body to form, even at their lower end of the z scale z=7.5 (0.71 Gyr)!
I'm afraid the standard model has been saved yet again.

hellfire
SpaceTiger said:
This issue was directly addressed by Jim Gunn back in 1972. See here.
Hey, cool, this reference and the one of the original paper of Press and Schechter were very nice. May I suggest you to sum up some "historical" references about cosmology and astrophysics an post them in the "A&C Reference Library". I am sure every one of us would be grateful.

SpaceTiger
Staff Emeritus
Gold Member
hellfire said:
May I suggest you to sum up some "historical" references about cosmology and astrophysics an post them in the "A&C Reference Library".
Sounds like a good idea. I'll see what I can do.

Garth
Gold Member
SpaceTiger said:
The key thing to notice here is that they're suggesting the stars formed at that epoch, not necessarily the galaxy. Remember, in a bottom-up scenario, small things form first.
Which part of
This implies formation epochs of zf ≈ 7.5 − 13.5 for the galaxies SBM03#1
do you not understand? (From that paper of observations of high z galaxies.) The stars were members of "a prominent older stellar population which probably dominates the stellar mass population" of these galaxies. (Italics my addition)
No, the reference above is simply about observations. I'm asking for a reference suggesting that there's a problem with CDM predicting these galaxies.
Yes, I was only referring to that paper, the model has to fit the observations, not the other way round. The baryonic gas in the early universe was pressurized, the pressure may have been low but so was the density, so the Jean's Mass and free fall time are relevant. Of course it is always possible to add in DM to make any model work, but until we know what that DM is the argument is hardly persuasive.

Garth

Last edited: