Can we really see 3D in the universe.

[arivero]

Our Earth orbit around the sun (and around the galaxy) can be considered a point when looking at other galaxies. So time ago I wondered how was that we assume we know a 3D-distribution of galaxies without having a real stereoscopic view of it.

I know, there is doppler. But if one galaxy is behind another, we will see only the doppler of the first one, will we?

 

[arivero]

dark matter

Hmm let me extend on the question. It was originated around 1995-96, when I heard a talk from Pietronero (or someone from his team) in Barcelona. A nice meeting, parting time with Mandelboot himself. I wondered if the fractal dimension D=2 was just a projection efect, assuming that galaxies have a non-null diameter and that we will have difficulties to see a galaxy that is hidden behind a nearer one. I am not sure if I got to show this point clearly at these times -I am afraid I didn't, althought I tryed some email contact-, but in the meantime a mathematical statement of this visibility problem has been done, see astro-ph/0301034

My question is, now, if this problem is real. I have seen some works on catalog bias, but I am not sure if they proof that we do not see all the galaxies, but only a 2-D fractal surface of the 3-D distribution.

Second question, and perhaps more important, is if this effect is accounted for, in the studies of dark matter. Does the studies of dark matter use the map of galaxy distribution? Because if it is biased towards 2-D, then for sure there is dark matter... just hidden behind the galaxies we see!

 

[chroot]

We can see Cepheid variable stars in many galaxies, which provide an excellent distance indicator. We use that to calibrate the Hubble redshift relation, which relates distance to recessional velocity. Then, we can use redshift to determine the distances to other galaxies.

I also think you're misusing the word "fractal." What we see on the sky is a 2D projection of a 3D world, not anything fractal. "Fractal" means having fractional (non-integer) dimension.

The bottom line is there is a preponderance of evidence, from observations of novae, star clusters, variable stars, and other objects within galaxies, that supports the widely-accepted distance estimates. Surely there is some systematic error in the process. In fact, there is an entire "ladder" of distance measurement techniques. Each more distance technique is calibrated against a nearer one. Thus, a change in the nearest technique (stellar parallax) can result in a change in the most distant technique. It is not without error.

All available evidence supports the model that some galaxies are close, while others are far away. The exact distances may be questionable, but the idea that they are at different distances really isn't.

- Warren

 

[Nereid]

To the question of 'seeing' one galaxy behind another ...

IIRC, there are some quite nice Hubble images which show just that - a distant galaxy or three seen between gaps in star lanes/spiral arms of a nearer one. How can we tell it's truly further away? Of course it depends how determined we are to find out! But should we want to, we could measure the redshifts of both galaxies, wait for a supernova in each, look for absorption lines from the nearer galaxy in the spectrum of the more distant one, ...

But what if the 'behind' galaxy were behind a spiral arm, or deeply hidden near the core of an elliptical? Well, there's always different EM bands; in the IR dust becomes much less opaque, in X-rays any neutron stars or the galactic nucleus will shine right through, in radio we might be lucky to find a bright source in the distant galaxy, ... In fact, the IR and radio are how we discovered several nearby galaxies (including some distant members of the Local Group) even though they were 'hidden' behind Milky Way clouds.

Second question, and perhaps more important, is if this effect is accounted for, in the studies of dark matter. Does the studies of dark matter use the map of galaxy distribution? Because if it is biased towards 2-D, then for sure there is dark matter... just hidden behind the galaxies we see!

Quick answer: yes, it is accounted for. One of the posts in the new A&C Sticky has a link to publications from the 2dF work; there are several papers there which discuss these effects.

 

[arivero]

Originally posted by chroot

I also think you're misusing the word "fractal." What we see on the sky is a 2D projection of a 3D world, not anything fractal. "Fractal" means having fractional (non-integer) dimension.

Ah no, by fractal I mean I am measuring the fractal dimension of a distribution of points. There is a lot of claims around telling that the fractal (content) dimension of galaxy catalogues is about d=2, instead of d=3. Now, this is the effect one expects if you blow out a 2D distribution, the celestial map, into a 3D space. It would indicate that we are not really seeing all the galaxies in the space, but only the nearest one in each line of sight.

Nereid, thanks for the pointer. I am going to check the sticky.

 

[russ_watters]

Originally posted by Nereid
To the question of 'seeing' one galaxy behind another ...

And a little more:
Since galaxies are so small relative to the field of view of the entire sky, its rare enough that they actually "block" each other.

 

[arivero]

Originally posted by russ_watters
And a little more:
Since galaxies are so small relative to the field of view of the entire sky, its rare enough that they actually "block" each other.

Hmm sure they are? It seems to me that the whole field of view is crowded with galaxies there at the end.

 

[Nereid]

cosmic wallpaper?

Originally posted by arivero
Hmm sure they are? It seems to me that the whole field of view is crowded with galaxies there at the end.

It's a fascinating idea - that if you look deep enough (limiting magnitude) every line of sight will hit a galaxy, with the most distant being the most likely.

IIRC, you can be quite confident of this (with some important caveats), provided your resolution limit is >~2". However, once you get to Hubble-type resolutions (~0.1"), or observe in different wavelengths, other 'wallpapers' become more important (e.g. in X-rays it's the hot IGM in clusters; for microwaves it's the CMB).

If you take a look at 100 (say) randomly chosen deep Hubble images, you'll get a good visual impression of cosmic wallpaper.

 

[russ_watters]

Originally posted by arivero
Hmm sure they are? It seems to me that the whole field of view is crowded with galaxies there at the end.

That's true: http://www.stsci.edu/ftp/science/hdf/hdf.html Still not a lot of overlap though.

 

[Nereid]

Originally posted by russ_watters
That's true: http://www.stsci.edu/ftp/science/hdf/hdf.html Still not a lot of overlap though.

Don't be fooled Russ; if the limiting magnitude of the HDF is 28 (say), and you can estimate what the galaxies would look like if you could see to 33, or 38 (that's only another 2 OOM fainter), or 48 (!), there'd be an awful lot less 'empty space' - check out the generalised theoretical profiles (e.g. King, NFW, de Vaucouleurs).

 

[arivero]

Can anyone provide a number? Ie if the total view is 4pi stereoradians, how many sradians are occuped by galaxies and how many are plain vacuum with nothing to see?

It seems I must give a look to Nereid pointers :-)

 

[GRQC]

Originally posted by arivero
Ah no, by fractal I mean I am measuring the fractal dimension of a distribution of points. There is a lot of claims around telling that the fractal (content) dimension of galaxy catalogues is about d=2, instead of d=3. Now, this is the effect one expects if you blow out a 2D distribution, the celestial map, into a 3D space. It would indicate that we are not really seeing all the galaxies in the space, but only the nearest one in each line of sight.

Nereid, thanks for the pointer. I am going to check the sticky.

We obtain 3-d maps of the (local) neighborhood via redshift surveys (SSRS, CfA, etc...) -- i.e. measure the observed redshift of the galaxies and use general relativistic expressions to calculate their distance (a good approximation is Mattig's formula).

There actually wasn't a lot of this data to go around until relatively recently, and the more recent analyses of these surveys started to again strengthed the notion of fractally-distributed matter (first suggested by Peebles in the late 70s or early 80s, I believe -- Charlier and deVaucouleurs notwithstanding!). This indicated that D=2 was the dominant spatial scaling power, as opposed D=3 as one would expect from the cosmological principle (homogeneity and isotropy). In fact, fractal distributions of matter violate the cosmological principle, indicating preferential spatial distributions.

The debate on this is still open, of course, but it nevertheless implies a interesting conundrum to be reconciled. Observationally, a D=2 scaling could imply sheet-like structures of galaxies, which are observed.

 

[russ_watters]

Originally posted by Nereid
Don't be fooled Russ; if the limiting magnitude of the HDF is 28 (say), and you can estimate what the galaxies would look like if you could see to 33, or 38 (that's only another 2 OOM fainter), or 48 (!), there'd be an awful lot less 'empty space' - check out the generalised theoretical profiles (e.g. King, NFW, de Vaucouleurs).

I'm looking forward to HDF2, due to be released this month.

 

[Nereid]

Let's do some back of the envelope work ourselves

Originally posted by arivero
Can anyone provide a number? Ie if the total view is 4pi stereoradians, how many sradians are occuped by galaxies and how many are plain vacuum with nothing to see?

I think we need to start with some definitions ... what do we mean when we say 'this galaxy occupies x arcsec²'?

Time was when the edge of a galaxy was the 25 mag/sq arcsec isophote (B band), often smoothed.

If we adopted that definition, we'd count many a LSB galaxy as empty space!

And, as we've been discussing, you can see through a galaxy; the recent work on the Andromeda halo (M31) - which is, incidently, the deepest image yet - is an impressive example:
http://hubblesite.org/newscenter/newsdesk/archive/releases/2003/15/

GRCQ wrote:*SNIP The debate on this is still open, of course, but it nevertheless implies a interesting conundrum to be reconciled. Observationally, a D=2 scaling could imply sheet-like structures of galaxies, which are observed.

Here's one of the more recent results (IIRC it nicely confirms and extends the 2dF work, which concluded last year): http://www.sdss.org/news/releases/20031028.powerspectrum.html