How close must a galaxy like our own be to clearly detect?

[TooMany]

Stated differently, with current instruments, at what distance would a galaxy like our own probably not be detected?

[Hornblower]

Stated differently, with current instruments, at what distance would a galaxy like our own probably not be detected?

With today's superpowered instruments we can detect them several billion light years away.

[TooMany]

Thanks. Can you be more precise? 3, 5, 10 billion light years? I know we have detected galaxies at over 13 billion light years, but I assume those have a much higher surface brightness or perhaps high output in the UV which gets red shifted.

[antoniseb]

If you'd be willing to take an answer to a similar question:
How far could M31 and M33 be and still be detected by our instruments, you might get a more precise set of numbers. One is larger than us, one is smaller, but both are within an order of magnitude. Both have been observed by Galex and other probes, so we have nice data about their UV surface brightnesses. We could probably answer that with some limited amount of work. The Milky Way's integrated brightness at various wavelengths might be harder to look up.

[TooMany]

Sure, that's close enough for the question I'm asking. So how far away could we reliably detect M31 and M33?

Another interesting question is how far away could we get spectral lines good enough to nail distance within say 1 GLY (or whatever is short for a billion light years).

It would be nice to have a better idea of our current limits.

[Nereid]

There's a very lengthy thread, in the Astronomy section, devoted to this question: Bias effects in galaxy detection.

Reader's digest summary: if you accept the (theoretical) analysis of Disney and Lang, out to z ~<1.2; if you accept the (observation-based) analysis of Fathi et al., out to z ~5.5 (and there's a curious - observation-based - paper near the end of the thread ).

[TooMany]

That's a big difference of opinion. I would trust observation over theory but then I guess the issue becomes exactly what is being observed.

Thanks for the link.

[TooMany]

There's a very lengthy thread, in the Astronomy section, devoted to this question: Bias effects in galaxy detection.

Reader's digest summary: if you accept the (theoretical) analysis of Disney and Lang, out to z ~<1.2; if you accept the (observation-based) analysis of Fathi et al., out to z ~5.5 (and there's a curious - observation-based - paper near the end of the thread ).

Wow, you are seriously trashing that D&L paper. Have you directly challenged the authors to respond to your criticisms?

I'm a bit left in the dark here. It's hard to understand how these estimates can be so far apart unless some important assumptions are different, such as how does Z relate to expected dimming.

[Jeff Root]

It's hard to understand how these estimates can be so far apart ...

I haven't looked at Nereid's thread. But when I read the first post
in this thread, my thought was that it probably couldn't be pinned
down to better than a factor of two. I'm the least knowledgeable
person on this subject to reply so far, though.

The Milky Way and Andromeda are both quite large, massive
galaxies, because they are old and are the result of mergers of
many smaller galaxies over billions of years. Any galaxies that
we can see at great distances will necessarily be younger. They
could be brighter, though, if they went through a period of intense
star formation while they were young.

Also, "nearby" distances -- millions of light-years -- are fairly
straightforward to compare. Cosmological distances -- billions
of light-years -- not so much. Comparing the brightness of a
galaxy six billion light-years away to that of one twelve billion
light-years away would likely be daunting for me.

-- Jeff, in Minneapolis

[Nereid]

Stated differently, with current instruments, at what distance would a galaxy like our own probably not be detected?

From your most recent post, I see that you have begun to read that thread. Probably you've concluded (if only temporarily) that to answer the question requires a few more assumptions. For example, do you include galaxy evolution (a galaxy observed to be many Gpc - gigaparsecs - distant will also be many billions of years younger)? Another: do you look for detection using every current instrument, across the entire electromagnetic spectrum (or just those in a narrow slice, say 400-700 nm)? One more: by "detect" do you mean something like "can obtain a reliable bulge plus disk intensity profile, out to ~10 kpc", or would something like "recognise the object as extended/bigger than the PSF"?

To take an extreme case: if a hypernova producing a long GRB happened in a galaxy like our own, that GRB could be detected to z ~7, maybe more.

A different extreme: galaxies like our own are hard to detect, with confidence, much beyond z ~0.3 in SDSS ("detect" here is deliberately ambiguous).

[TooMany]

From your most recent post, I see that you have begun to read that thread. Probably you've concluded (if only temporarily) that to answer the question requires a few more assumptions. For example, do you include galaxy evolution (a galaxy observed to be many Gpc - gigaparsecs - distant will also be many billions of years younger)?

I don't want to consider evolution. What I'm actually curious about is if there are (were) galaxies similar to our own at the time that they emitted the light reaching us now, at what distance would we be able to detect them?

Another: do you look for detection using every current instrument, across the entire electromagnetic spectrum (or just those in a narrow slice, say 400-700 nm)? One more: by "detect" do you mean something like "can obtain a reliable bulge plus disk intensity profile, out to ~10 kpc", or would something like "recognise the object as extended/bigger than the PSF"?

To take an extreme case: if a hypernova producing a long GRB happened in a galaxy like our own, that GRB could be detected to z ~7, maybe more.

A different extreme: galaxies like our own are hard to detect, with confidence, much beyond z ~0.3 in SDSS ("detect" here is deliberately ambiguous).

You make a good point that the meaning of detect is not well defined. Without attempting to get very precise what I mean is at what distance would galaxies like our own be "seen" to the extent that with some confidence we could say, yes there's definitely a galaxy there and at the time the light that we detect was emitted, that galaxy could have been similar to our galaxy as it is now.

I'm really focusing on an issue similar to that discussed in the thread you pointed out. Yes we detect some sort of galaxies at extreme distances but those are not likely similar to our own as it currently is when the light was emitted. I.e. they were much more luminous.

Let's put it another way, if there were lots of galaxies like our own at the time they emitted the light that reaches us now, at what maximum distance would we be well aware of their existence. I realize this is a theoretical question because of the assumptions about dimming and dust, so I'm asking in the context of mainstream LCMD theory formulas. So far I've got a range from 0.3 to 5.5 Z. That actually isn't as bad as it sounds since (using Dr. Wright's calculator) it's roughly 3.4Gly to 12.6Gly, less than a factor of 4 in range. I'm a bit skeptical about the 5.5 Z figure which is close to the BB. How did you come up with that limit?

Isn't Z a rather odd choice of measures? It is not linear with distance or time. Perhaps Gly or fraction of age of the universe we be easier to grasp. (I guess one good thing about Z is that it's an actual measure of something, not derived from theory.)

Finally here's another perhaps similar question. Is there a distance at which we are sure that we could detect galaxies like our own (as it currently is) but we do not, meaning that such galaxies did not yet exist at that time.

[antoniseb]

... So far I've got a range from 0.3 to 5.5 Z. That actually isn't as bad as it sounds since (using Dr. Wright's calculator) it's roughly 3.4Gly to 12.6Gly, less than a factor of 4 in range. ...

The z~0.3 is an estimate for looking face-on in images from the SDSS. The SDSS is a survey with what is now a medium-sized telescope BUT it is looking at a large fraction of the sky with relatively short exposures, and catches lots of galaxies. Your question seemed more to do with if there were lots of such mature galaxies, how far away could we see them by chance in deep images such as the Hubble Ultra Deep Field, or deep images collected in the near infrared by the VLT. That z value would probably be larger than 1.0 and might be closer to the 4 or 5 we've seen above, but I don't have a figure to give you for it.

[Nereid]

Wow, you are seriously trashing that D&L paper. Have you directly challenged the authors to respond to your criticisms?

No.

But then I don't think I have to, considering ... well, let me just say that, when you get to the last page of that thread (p8), I reference an interesting paper.

I'm a bit left in the dark here. It's hard to understand how these estimates can be so far apart unless some important assumptions are different,

Yes, the assumptions used are very important. That's partly why I asked you to clarify your question (in the OP).

such as how does Z relate to expected dimming.

What Disney and Lang do (or rather, claim to do) is put a galaxy somewhat like our own (the Milky Way) at various distances (redshifts) and estimate whether it could be detected or not. They use two kinds of "detection" - could the galaxy be 'seen' as an extended source? and could the galaxy be 'seen' as a point source?

Redshift (z) affects the former via observed surface brightness (which goes as (1+z)^4) ), while how the latter varies with z is a bit more complicated.

[Nereid]

I don't want to consider evolution. What I'm actually curious about is if there are (were) galaxies similar to our own at the time that they emitted the light reaching us now, at what distance would we be able to detect them?

OK, but I think that's a highly restrictive assumption.

Consider AGNs (active galactic nuclei): our galaxy does not, today, have an AGN; however, per our current understanding, it certainly did in the past, possibly as recently as a few million years ago. And its nucleus certainly was active many times before that*. Nearly a decade ago now Chandra stared at a small field in the constellation Boötes for a while, and found lots of point sources (and a few extended sources). Many of these turned out to be AGN, some at very high redshift. Optical counterparts for some of the x-ray sources were not found. Is it possible that one of such sources may have been a distant Milky Way, with its SMBH+accretion disk 'on'?

Now suppose an x-ray survey is done, with Chandra staring at a small field for much longer. For sure, many more point sources will be detected (some caveats apply), and perhaps some will be just like our galaxy, at a considerably greater distance. As luck would have it, here is a recent report of just such a deeper x-ray survey.

You make a good point that the meaning of detect is not well defined. Without attempting to get very precise what I mean is at what distance would galaxies like our own be "seen" to the extent that with some confidence we could say, yes there's definitely a galaxy there and at the time the light that we detect was emitted, that galaxy could have been similar to our galaxy as it is now.

That's a lot more precise, and so the answer to your question can be too.

Of course, you say "with some confidence" but also "definitely"!

You also say "light"; do you mean "electromagnetic radiation, in the region ~400-700 nm" (i.e. the 'visual' or 'optical')? Or do you mean any part, or parts, of the entire electromagnetic spectrum?

In any case, as far as I know, it's not possible to estimate the distance to a dim, extra-galactic object - whether point source or extended source - unless you can reliably estimate its redshift. A good spectrum - anywhere in the UV to MIR - would give you that, with a high degree of confidence; you can also get a 'photo-z' (i.e. an estimate of redshift based on photometry), but it has a considerably lower confidence, especially for an individual object.

So it's relatively easy to detect ('see') a galaxy, but not so easy to estimate, with confidence, its redshift.

I'm really focusing on an issue similar to that discussed in the thread you pointed out. Yes we detect some sort of galaxies at extreme distances but those are not likely similar to our own as it currently is when the light was emitted. I.e. they were much more luminous.

The key point the Disney and Lang paper makes is not so much the luminosity (brightness) one, but the surface brightness one: at extreme distances, our galaxy's apparent surface brightness is too low for it to be detected.

Let's put it another way, if there were lots of galaxies like our own at the time they emitted the light that reaches us now, at what maximum distance would we be well aware of their existence. I realize this is a theoretical question because of the assumptions about dimming and dust, so I'm asking in the context of mainstream LCMD theory formulas. So far I've got a range from 0.3 to 5.5 Z. That actually isn't as bad as it sounds since (using Dr. Wright's calculator) it's roughly 3.4Gly to 12.6Gly, less than a factor of 4 in range. I'm a bit skeptical about the 5.5 Z figure which is close to the BB. How did you come up with that limit?

Isn't Z a rather odd choice of measures? It is not linear with distance or time. Perhaps Gly or fraction of age of the universe we be easier to grasp. (I guess one good thing about Z is that it's an actual measure of something, not derived from theory.)

Finally here's another perhaps similar question. Is there a distance at which we are sure that we could detect galaxies like our own (as it currently is) but we do not, meaning that such galaxies did not yet exist at that time.

I'll address this part of your post later; stay tuned!

* true, the 'duty cycle' - as it is called - of the SMBH (super-massive black hole), and its accretion disk, at the centre of the nucleus of galaxies like ours is not well understood, but they do seem to have been 'on' more than once in their lives

[TooMany]

Redshift (z) affects the former via observed surface brightness (which goes as (1+z)^4) ), while how the latter varies with z is a bit more complicated.

That makes the ratio of surface brightness for 0.3 to 5.5 Z about 625, nearly three orders of magnitude.

The brightness ratio of the D&L estimate for local-like galaxy disappearance of 1.2 Z and your estimate of 5.5 Z is 92 or almost two orders of magnitude using that equation. That's a pretty big disagreement!

Interesting that surface brightness just goes crazy low with increasing Z, even though the difference between say 5 Z and 10 Z in terms of distance is not much (less than 1billion light years), the dimming is a factor of 11 but the difference in distance is only 10%. It's a little hard to get my head around how the dimming can become so radical just "slightly" further away. I suppose Z goes to infinity at BB T0? I must have some misconception about the distances in an expanding Universe or something. In Ned Wright's calculator I'm taking the light travel time as the measure of distance.

1 + Z is the wavelength ratio (emitted/received) of a given spectral line, correct? Strange that Z was defined that way, instead of simply as the ratio.

[TooMany]

As luck would have it, here is a recent report of just such a deeper x-ray survey.

Interesting but I found it hard to gather the import of paper except that they found that where only AGN types could be detected before they now can detect normal galaxies via soft X-rays and the number of normal galaxies detected is approaching the number of AGN types detected. Not sure if we are supposed to be able to conclude much new from that, except perhaps that at half the current age, there were at least as many non-active as active galaxies. The HST optical images clearly show "normal" galaxies at 1 Z, but I have no idea if those galaxies are comparable to our own in luminosity.

BTW I guess the X-ray detections of AGN are relatively insensitive to the orientation of the jets? Otherwise wouldn't it be possible that the "normal" galaxies detected are actually AGN at unfavorable viewing angles?

In any case, as far as I know, it's not possible to estimate the distance to a dim, extra-galactic object - whether point source or extended source - unless you can reliably estimate its redshift. A good spectrum - anywhere in the UV to MIR - would give you that, with a high degree of confidence; you can also get a 'photo-z' (i.e. an estimate of redshift based on photometry), but it has a considerably lower confidence, especially for an individual object.

So it's relatively easy to detect ('see') a galaxy, but not so easy to estimate, with confidence, its redshift.

OK, so let's make it a bit more precise. Detection will be defined to mean that the source can be resolved as extended and that a spectral line can be measured to get Z. That's what I would like to consider as confident identification.

[Nereid]

Continuing ...

Let's put it another way, if there were lots of galaxies like our own at the time they emitted the light that reaches us now, at what maximum distance would we be well aware of their existence.

In a later post, you moved a very long way from this as the key criterion!

As you discovered, 'being aware of the existence' of something, and being able to convince yourself of its similarity to something else, are two very different questions (or goals).

The purpose - well, one purpose - of my 'x-ray papers and post' was to illustrate that you could very well 'be aware of' a galaxy much like a (somewhat) earlier/younger version of our galaxy (i.e. one with an active nucleus) without knowing, with any degree of certainty, that it is, in fact, a Milky Way sibling.

I realize this is a theoretical question because of the assumptions about dimming and dust, so I'm asking in the context of mainstream LCMD theory formulas.

I don't know how dust entered your considerations; perhaps you were concerned that a galaxy like ours, viewed sideways (so that the disk appears edge-on) might be very much dimmer (in that narrow band, ~400-700 nm) than the same galaxy viewed face on.

Can you explain why you thought this important, please?

So far I've got a range from 0.3 to 5.5 Z. That actually isn't as bad as it sounds since (using Dr. Wright's calculator) it's roughly 3.4Gly to 12.6Gly, less than a factor of 4 in range. I'm a bit skeptical about the 5.5 Z figure which is close to the BB. How did you come up with that limit?

It's from the Fathi et al. paper I indirectly mentioned (I did some rounding), in an earlier post; have you read it yet?

Oh, and is there are particular reason you write "Z" for redshift, instead of the usual convention ("z")?

Isn't Z a rather odd choice of measures? It is not linear with distance or time. Perhaps Gly or fraction of age of the universe we be easier to grasp. (I guess one good thing about Z is that it's an actual measure of something, not derived from theory.)

It's what you can readily observe; converting it to a distance depends on the cosmological model you choose to use, and the values of the parameters in that model.

It also depends upon the definition of distance you're using, e.g. luminosity distance, radar distance, etc. Which one have you chosen?

Finally here's another perhaps similar question. Is there a distance at which we are sure that we could detect galaxies like our own (as it currently is) but we do not, meaning that such galaxies did not yet exist at that time.

The answer to that question is a definite "maybe"!

It depends, heavily, on things like what you mean by "galaxies like our own (as it currently is)" and "detect".

But we've moved on since your wrote these words; time for me to do so too.

[TooMany]

I mentioned dust in reference to intergalactic dust which affects brightness and color of background galaxies. Z versus z - oops, but you got it anyway.

It seems that in this era of "precision cosmology" one should take conclusions about the galactic population at z ~> 0.15 with a large grain of salt.

I do wish the astronomical community would be more modest about what they think they know.

[TooMany]

It's from the Fathi et al. paper I indirectly mentioned (I did some rounding), in an earlier post; have you read it yet?

Yes I read it. So what accounts for the itty-bitty galaxies at high Z; bizarre early galaxies or the wrong formulas for angular size and surface brightness?

I think that the "as much as 8 times smaller" reported may be consistent with no-expansion where z is the result of a fractional decline of energy proportional to distance.

[Nereid]

That makes the ratio of surface brightness for 0.3 to 5.5 Z about 625, nearly three orders of magnitude.

The brightness ratio of the D&L estimate for local-like galaxy disappearance of 1.2 Z and your estimate of 5.5 Z is 92 or almost two orders of magnitude using that equation. That's a pretty big disagreement!

Yes it is.

And, given how much of that other thread you've no doubt read by now, you have begun to appreciate some of the limitations of D&L's approach. Yes?

Interesting that surface brightness just goes crazy low with increasing Z, even though the difference between say 5 Z and 10 Z in terms of distance is not much (less than 1billion light years), the dimming is a factor of 11 but the difference in distance is only 10%. It's a little hard to get my head around how the dimming can become so radical just "slightly" further away.

Quite astonishing how General Relativity works, isn't it?

Yet it's a very well tested theory, with nothing even remotely in sight to challenge it.

I suppose Z goes to infinity at BB T0?

You have, I'm sure, heard of the CMB, the Cosmic Microwave Background?

Do you know what the estimated z for that is? And how long, after t=0, it is estimated to have occurred?

I must have some misconception about the distances in an expanding Universe or something. In Ned Wright's calculator I'm taking the light travel time as the measure of distance.

Why choose this kind of distance, rather than any other?

1 + Z is the wavelength ratio (emitted/received) of a given spectral line, correct? Strange that Z was defined that way, instead of simply as the ratio.

Well, physics is full of strange conventions. And astronomers have them too ... did you know that, to an astronomer, oxygen is a metal? as is carbon, neon, nitrogen, ... in fact every element is a metal, to an astronomer, other than hydrogen and helium (although I'm not sure lithium is, to every astronomer, always a metal).

[Nereid]

Interesting but I found it hard to gather the import of paper except that they found that where only AGN types could be detected before they now can detect normal galaxies via soft X-rays and the number of normal galaxies detected is approaching the number of AGN types detected. Not sure if we are supposed to be able to conclude much new from that, except perhaps that at half the current age, there were at least as many non-active as active galaxies.

Here's a secret: I've been trying to get you to reveal your 'visual chauvinism'!

How? By finding an example of the detection of a galaxy 'just like' our own, but only in a part of the electromagnetic spectrum that you're (likely) unfamiliar with. Then, if I could show that such a detected galaxy is at z~5 (say), the question you posed in the OP would be answered in a way that was, I expect, wholly surprising and unexpected.

But you spoiled the fun, by changing your criteria before I could get here.

Never mind, what about this: if a Milky Way look-alike were 'seen', in the 'light' of 21-cm HI, at a distance of z~2, would that count as a (partial) answer to your original question? If not, why not?

The HST optical images clearly show "normal" galaxies at 1 Z, but I have no idea if those galaxies are comparable to our own in luminosity.

Do you know what wavelength those HST images were taken in? And if the HST were somewhere in M31, looking at our own galaxy, what wavelength would the HST have to be observing at in order to take comparable images?

More generally, how do you convert the observed luminosity (brightness) of a z ~1 galaxy in an HST image into its estimated total luminosity if it were concentrated into a point, located 10 pc from us? If you don't know, just say so; I'll be more than happy to explain (HINT: the answer is in that 8-page thread).

BTW I guess the X-ray detections of AGN are relatively insensitive to the orientation of the jets? Otherwise wouldn't it be possible that the "normal" galaxies detected are actually AGN at unfavorable viewing angles?

As I understand it, spiral galaxy AGNs rarely, if ever, have jets. Certainly not ones visible in the x-ray part of the spectrum.

The source of x-rays, in most AGNs, is the accretion disk, and the immediate surroundings (by reflection). In principle, you'd expect accretion disk orientation - relative to us - would matter a lot (in terms of observed x-ray intensity); I do not know if, in fact, it does.

OK, so let's make it a bit more precise. Detection will be defined to mean that the source can be resolved as extended

Good.

But here comes a possible problem: for a source (object) which is just barely resolved, how do you tell if it's a spiral (and not an elliptical)? And even if you can tell it's a spiral (by some magic), how can you tell it's a spiral like the Milky Way? In the local group, there are three classical spirals: M31, our galaxy, and M33. They differ considerably in size and estimated total luminosity. They also differ - considerably - in many other ways (e.g. M31 apparently has far more globular clusters than our galaxy does, and M33's globular clusters seem to have a very different range of ages than those of either M31 or our own galaxy).

and that a spectral line can be measured to get Z.

Emission or absorption? Or you don't care?

And if there's only one, how do you know which one it is? If you make a mistake, you'll get the estimated z completely wrong, won't you?

That's what I would like to consider as confident identification.

Assuming the redshift is confident (by some magic), wouldn't that simply result in you having identified an extended source, of redshift z? How would you know it's even a galaxy?

[TooMany]

... you've no doubt read by now, you have begun to appreciate some of the limitations of D&L's approach. Yes?

Well I would say their analysis is simplistic. But, as you point out, a statement like "the local group would dropout viewed from z=1.2" all depends on what counts as visibility of the local group. Certainly at that distance our ability to measure the properties of galaxies is substantially affected in comparison with nearby galaxies. Surface brightness has fallen to 1/23 of local (if I've got the right formula). Thus the reliability of conclusions about galactic evolution is rather weak which I believe that is the whole point of the D&L paper.

Quite astonishing how General Relativity works, isn't it?
Yet it's a very well tested theory, with nothing even remotely in sight to challenge it.

Relative to our Euclidean intuition, GR is truly bizarre.
I think you forget that the failure to find the DM is itself a rather serious challenge to Einstein's formulation for gravity and that there are alternatives being proposed and refined.

You have, I'm sure, heard of the CMB, the Cosmic Microwave Background?

Do you know what the estimated z for that is? And how long, after t=0, it is estimated to have occurred?

I'm aware of these theoretical conclusions - z~1000 and t~380Ky IIRC.

Why choose this kind of distance, rather than any other?

To get a sense of how far the light had to travel to reach us. Is it a poor choice? What do you recommend? The comoving distance isn't particularly interesting in this discussion.

Well, physics is full of strange conventions. And astronomers have them too ... did you know that, to an astronomer, oxygen is a metal? as is carbon, neon, nitrogen, ... in fact every element is a metal, to an astronomer, other than hydrogen and helium (although I'm not sure lithium is, to every astronomer, always a metal).

I'm not a trained astronomer nor even a physicist, yet anyone who has dabbled in astronomy knows this convention. Since you brought it up, what's your take on observations of metallicity of very distant galaxies? I think I read recently that the farthest quasar? yet discovered was found to be chock-full of carbon.

[TooMany]

Here's a secret: I've been trying to get you to reveal your 'visual chauvinism'!

How? By finding an example of the detection of a galaxy 'just like' our own, but only in a part of the electromagnetic spectrum that you're (likely) unfamiliar with. Then, if I could show that such a detected galaxy is at z~5 (say), the question you posed in the OP would be answered in a way that was, I expect, wholly surprising and unexpected.

We'll we can detect GRB very, very far, but that provides little information about the nature of the source galaxies. You may detect some red-shifted UV, but it is very hard to know from that what sort of object you are seeing. The further galaxies are away, the less we can know about them and the more speculative conclusions become. Not only because of dimming and angular size, but also because the interpretation of what we "see" becomes more and more driven by theory and less by direct observation.

Never mind, what about this: if a Milky Way look-alike were 'seen', in the 'light' of 21-cm HI, at a distance of z~2, would that count as a (partial) answer to your original question? If not, why not?

Hard to say without knowing what we can discern using 21-cm at z~2. For example, could we resolve the spiral arms. Could we make corroborating measurements at other wavelengths?

Do you know what wavelength those HST images were taken in? And if the HST were somewhere in M31, looking at our own galaxy, what wavelength would the HST have to be observing at in order to take comparable images?

The images vary from z~0.08 to 1.2. All were made with what I'm guessing are visible band filters (B435, V606, and z850). I cannot find the absolute magnitudes in the paper. One at z 1.2 is quite red and dim but looks to be a face on spiral. The absolute mag of M31 is about -20. So now, I do wonder what it would look like (imaged identically) at z~1.2!

More generally, how do you convert the observed luminosity (brightness) of a z ~1 galaxy in an HST image into its estimated total luminosity if it were concentrated into a point, located 10 pc from us? If you don't know, just say so; I'll be more than happy to explain (HINT: the answer is in that 8-page thread).

No offense please, but why should I care? Isn't surface brightness more useful to talk about. I would imagine that compressing the luminosity into a point would much improve visibility/detectability.

As I understand it, spiral galaxy AGNs rarely, if ever, have jets. Certainly not ones visible in the x-ray part of the spectrum.

That's interesting and something I was not aware of. What sort of galaxies are found with jets then? Are AGNs and quasars thought to be the same phenomenon, the result of central SMBHs in galaxies?

But here comes a possible problem: for a source (object) which is just barely resolved, how do you tell if it's a spiral (and not an elliptical)? And even if you can tell it's a spiral (by some magic), how can you tell it's a spiral like the Milky Way?

In this, I think we are in much agreement. The various sorts of detection discussed don't necessary tell us that much detail about what we have detected at cosmic distances. Again I think this is the whole point of the D&L paper. I'm not sure I see the basis for your severe criticism. I guess I could read that thread again but it really lost me in the details.

Emission or absorption? Or you don't care?

Definitely emission, who knows where the absorbing medium is relative to the emitter? Of course you could also argue about re-emission but that seems much less likely to be a problem.

And if there's only one, how do you know which one it is? If you make a mistake, you'll get the estimated z completely wrong, won't you?

You are a stickler, but of course you would need many more lines to identify the elemental source of any measured line.

Assuming the redshift is confident (by some magic), wouldn't that simply result in you having identified an extended source, of redshift z? How would you know it's even a galaxy?

You are continuing to make the D&L paper's point! Why did you get so upset about it?
(I wasn't aware that magic was required to be confident about a red-shift. Are you referring to my failure to mention the need for multiple lines?)

I'm very interested in what you can tell me about the above very specific question about an HST image of M31 at z~1.2.

Thanks!

[Nereid]

General readers may be interested in some images, to get an idea of what spiral galaxies look like.

Many BAUT members are - I hope! - aware of the Zooniverse, and its first Citizen Science project, Galaxy Zoo. Further, I hope that you have all signed up, and spent many happy hours as citizen scientists, classifying galaxies, studying galaxy mergers, or solar storms, discovering supernovae, finding new exoplanets, or plotting bubbles in space.

I certainly have, and I also spend an enjoyable few minutes (or more) every day reading some of the blogs or posts on the forums associated with each of these projects. Directly relevant to this thread is the galaxy zoo forum, and one post in particular: two classical spirals; one near, one far*.

From only the SDSS image of this z=0.239 spiral galaxy, you'd be hard pressed to say if this was a Milky Way look-alike or not. However, from the Hubble image you can instantly see that, while it's a spiral, it's certainly not at all like the Milky Way**!

At greater distances what do spiral galaxies look like? Well, even with the Hubble, not many look like spirals, but many are clearly disk galaxies. For example, look at the postage-stamp images in this recent paper: CANDELS: Correlations of SEDs and Morphologies with Star-formation Status for Massive Galaxies at z ~ 2.

* I tried to post the two images, but got a strange error message, so you'll have to click on the link to see how dramatically different the two images, of the same galaxy, are.
** it is obvious to you, isn't it?

[Nereid]

I mentioned dust in reference to intergalactic dust which affects brightness and color of background galaxies.

As far as I know, there are very few reports of the confident detection of intergalactic dust ... once you get out to ~5 effective radii for most spirals, the sky is quite clear. There are some exceptions - near the centre of rich clusters, possibly in various streams, and in tidal tails (a result of close encounters of the galactic kind). The IGM (inter-galactic medium) does certainly contain metals, but none of it in the form of dust.

Perhaps you were thinking of the ISM (interstellar medium)? If not, could you give some references please?

Z versus z - oops, but you got it anyway.

It seems that in this era of "precision cosmology" one should take conclusions about the galactic population at z ~> 0.15 with a large grain of salt.

One should always take broad, over-simplified conclusions with a truckload of salt!

Galaxies are fascinating astronomical objects; on the one hand they seem so simple, yet when you start to examine them in detail, they turn out to be amazingly complicated. Fortunately, astronomers being (in general) especially fond of paying a lot of attention to detail, conclusions they draw (in the papers they publish) come with more than enough detail to enable you, the reader, to estimate just how much salt you need to take.

I do wish the astronomical community would be more modest about what they think they know.

In what way (in particular, with relevance to this thread)?

[Nereid]

It's from the Fathi et al. paper I indirectly mentioned (I did some rounding), in an earlier post; have you read it yet?

Yes I read it. So what accounts for the itty-bitty galaxies at high Z; bizarre early galaxies or the wrong formulas for angular size and surface brightness?

I think that the "as much as 8 times smaller" reported may be consistent with no-expansion where z is the result of a fractional decline of energy proportional to distance.

I'll park this, for now, and come back to it later ...

[Nereid]

Most of this post is scene-setting, making sure we're on the same page, that sort of thing ...

... you've no doubt read by now, you have begun to appreciate some of the limitations of D&L's approach. Yes?

Well I would say their analysis is simplistic. But, as you point out, a statement like "the local group would dropout viewed from z=1.2" all depends on what counts as visibility of the local group. Certainly at that distance our ability to measure the properties of galaxies is substantially affected in comparison with nearby galaxies. Surface brightness has fallen to 1/23 of local (if I've got the right formula). Thus the reliability of conclusions about galactic evolution is rather weak which I believe that is the whole point of the D&L paper.

Good, especially the part about making doubly sure you're comparing apples with applies (so to speak).

Quite astonishing how General Relativity works, isn't it?
Yet it's a very well tested theory, with nothing even remotely in sight to challenge it.

Relative to our Euclidean intuition, GR is truly bizarre.
I think you forget that the failure to find the DM is itself a rather serious challenge to Einstein's formulation for gravity and that there are alternatives being proposed and refined.

Discussion of this would take us too far from this thread, but if DM poses a problem for GR, it poses an equally daunting problem for the weak field limit we call Newtonian gravity (e.g. spiral galaxy rotation curves).

You have, I'm sure, heard of the CMB, the Cosmic Microwave Background?

Do you know what the estimated z for that is? And how long, after t=0, it is estimated to have occurred?

I'm aware of these theoretical conclusions - z~1000 and t~380Ky IIRC.

Thus reinforcing the point that converting an observation of a redshift to an estimate of a distance is 'model dependent', as they say ...

Why choose this kind of distance, rather than any other?

To get a sense of how far the light had to travel to reach us. Is it a poor choice? What do you recommend? The comoving distance isn't particularly interesting in this discussion.

I think it depends on what you're trying to do, with the question you asked.

If your aim has something to do with galaxy evolution, then perhaps the best distance to use would be one that you could easily use to relate observations to 'age since photons streamed free' (i.e. when the universe ceased being radiation dominated). But as you have stated that you are explicitly interested in what the Milky Way would look like if it were a distance x from us, then perhaps a better measuring rod would be redshift (z)?

Well, physics is full of strange conventions. And astronomers have them too ... did you know that, to an astronomer, oxygen is a metal? as is carbon, neon, nitrogen, ... in fact every element is a metal, to an astronomer, other than hydrogen and helium (although I'm not sure lithium is, to every astronomer, always a metal).

I'm not a trained astronomer nor even a physicist, yet anyone who has dabbled in astronomy knows this convention. Since you brought it up, what's your take on observations of metallicity of very distant galaxies? I think I read recently that the farthest quasar? yet discovered was found to be chock-full of carbon.

Good; my point was that conventions sometimes do not seem to make much sense, when you are new to a field of science.

Let's discuss the estimated metallicity of high-z quasars in a separate thread, shall we?

[Nereid]

Here's a secret: I've been trying to get you to reveal your 'visual chauvinism'!

How? By finding an example of the detection of a galaxy 'just like' our own, but only in a part of the electromagnetic spectrum that you're (likely) unfamiliar with. Then, if I could show that such a detected galaxy is at z~5 (say), the question you posed in the OP would be answered in a way that was, I expect, wholly surprising and unexpected.

We'll we can detect GRB very, very far, but that provides little information about the nature of the source galaxies. You may detect some red-shifted UV, but it is very hard to know from that what sort of object you are seeing. The further galaxies are away, the less we can know about them and the more speculative conclusions become. Not only because of dimming and angular size, but also because the interpretation of what we "see" becomes more and more driven by theory and less by direct observation.

Which is getting us much closer to what it is, exactly, that you think makes the Milky Way unique.

And here we run into a completely different kind of challenge: what, exactly, does our own galaxy look like? As we're at a location close to the disk plane, ~10 kpc from the nucleus, it's actually very difficult to say, with a reasonable degree of confidence, what our homegalaxy looks like, from M31, say.

To get around this, how about we change the question in the OP to "How close must a galaxy like M31 be to clearly detect?" and "Stated differently, with current instruments, at what distance would a galaxy like M31 probably not be detected?"

If we use a GRB as a back-light, then detecting an M31 look-alike could be as simple as finding a set of absorption lines in its spectrum, at a fixed z, and with the characteristics of M31's ISM (a certain metallicity, and temperature). A bright quasar could serve in place of a GRB, giving the added benefit of repeat observations using different instruments over many years. In this way, an M31 look-alike could be detected to a redshift just shy of that of the GRB or quasar.

In emission, we could add up the light detected from the distant M31 look-alike, across as many wavelengths as possible*, and plot the integrated intensity against wavelength (or frequency). This would give us a SED (spectral energy distribution, or density); if we have a z, we can convert this to absolute units (assuming isotropic emission). We could then compare that with the (suitably redshifted) observed SED of M31 (taking care to ensure we integrate over the entire observed extent of M31, which is far from trivial to do). For this we do not need to 'resolve' the distant galaxy, although having it "... be detected in minimally four connected pixels, each at ≥ 0.75σ above the local computed sky-background" would be helpful**.

Never mind, what about this: if a Milky Way look-alike were 'seen', in the 'light' of 21-cm HI, at a distance of z~2, would that count as a (partial) answer to your original question? If not, why not?

Hard to say without knowing what we can discern using 21-cm at z~2. For example, could we resolve the spiral arms. Could we make corroborating measurements at other wavelengths?

OK, so this is worth exploring further then.

On the question of what "current instruments" (per the OP) are capable of, is it OK to include what facilities such as LOFAR and the SKA will be capable of, once they are up and running (and assuming they achieve their design objectives)?

* as the distant galaxy will be faint, the points may have to be integrated over a 'broad band', such as SDSS' r.
** see p8 of the Bias effects in galaxy detection thread.

(to be continued)

[TooMany]

Directly relevant to this thread is the galaxy zoo forum, and one post in particular: two classical spirals; one near, one far*.

From only the SDSS image of this z=0.239 spiral galaxy, you'd be hard pressed to say if this was a Milky Way look-alike or not. However, from the Hubble image you can instantly see that, while it's a spiral, it's certainly not at all like the Milky Way**!

I recall some years ago HST images of deep objects were interpreted as irregular, unformed galaxies. I think this interpretation was later discarded. The above example demonstrates how form can be misinterpreted in faint images.

Are you referring to the absence of a bar in that galaxy?

At greater distances what do spiral galaxies look like? Well, even with the Hubble, not many look like spirals, but many are clearly disk galaxies. For example, look at the postage-stamp images in this recent paper: CANDELS: Correlations of SEDs and Morphologies with Star-formation Status for Massive Galaxies at z ~ 2.

That helps answer my question(s). Interesting that the authors don't highlight how this population might substantially differ from the current population. The featured conclusion is that the merger theory of large galaxy formation is probably wrong. However they do mention star formation rates as much higher. This makes me wonder how star formation rates can be measured.

[ngc3314]

Interesting that the authors don't highlight how this population might substantially differ from the current population.

It's not my experience that much useful can be inferred from whether an individual research paper stresses or even mentions something or not. This is even more the case for something like CANDELS where the results will be appearing over perhaps dozens of papers over a span of years.

This makes me wonder how star formation rates can be measured.

Optical and UV color, luminosity in emission lines (especially recombination lines), radio emission... it seems that the (emitted) optical and near-infrared bands are the only ones in which we easily see stellar populations other than quite young ones. More details.