Dear Nereid and Phobos (seat belt on) can we talk about redshifts?

[Nereid]

How big are the public GOODS* data files? Seat belts on please (just the HST BViz observations) ... " The files for each section, both the science and the weight map images, are ~268.4 MB in size each (8192 x 8192 pixels, each with a 32 depth), and there are images in each of the four GOODS passbands. Thus, the CDF-S data set requires 2x268.4x18x4 = 38.655 GB, the HDF-N requires 2x268.4x17x4 = 36.507 GB. The full data set requires 38.655 + 36.507 = 75.162 GB."

*"GOODS will survey approximately 300 square arcmin divided into two fields: the Hubble Deep Field North and the Chandra Deep Field South. These are among the most data-rich portions of the sky, and are the sites of the deepest observations from Hubble, Chandra, ESA's XMM-Newton, and from many ground-based facilities. Dividing our survey area provides insurance against cosmic variance due to galaxy clustering, and guarantees that astronomers in both hemispheres can carry out related observations." ... and Spitzer too.

 

[big-egg]

Questions?

Are the disparate red shifts associated only with quasars and galaxies?

Are there coupling galaxies with disparate red shifts?

Are there supernovas in the proximity of the quasars?

What about the quantum red shifts? They will also create a hard problem to the BBT.

Are there statistical steadies on disparate red shifts?

Thank you for your time to consider these thought provoking questions.

 

[turbo]

Thank you Chronos

I also think Arp, while well intended, is statistically challenged and has a history of using dubious math to arrive at equally dubious conclusions: which is why he does not have much mainstream support. A good example is the Newman-Terzian paper http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=1995ApJ...441..505N&db_key=AST&high=3325b47acc06425

I am very surprised that you cited that statistically flawed paper, but I am very happy that you did so. The authors misstated Arp's premise and then applied statistics in a very creative way to downplay the improbability of something truly implausible.

Let's say you have a large galaxy and eleven associated smaller galaxies, clusters, whatever. If the smaller objects are physically associated with (and effected by) the more massive object, we would reasonably expect about half the smaller objects moving around that massive galaxy would be moving toward us relative to the host, and would therefore be blueshifted in relation to the host. If the host has one object it around it, there is one chance in two (1/2) that the object would be redshifted with respect to the host. If there are two objects, there is one chance in four (1/4) that both objects would be redshifted relative to the host. To save time, we will extrapolate: if there are eleven objects around the host, there is only one chance in 4096 that all eleven objects will be found to be redshifted relative to the host.

In the paper cited above, the authors de-coupled the massive host galaxy from any effects on the associated small objects, and then restated the problem as a simple matter of ordination, saying in effect "there is one in twelve chances for the host galaxy to have the smallest redshift so that's a 1/12 chance for the observation that all the smaller objects will be redshifted relative to the large massive galaxy." That has to be one of the most cynical applications of "statistics" I have seen.

The problem that Arp was illustrating is that lower-mass companions of large galaxies are preferentially redshifted with respect to their hosts (like M51 and NGC 5195). It is easy to statistically refute the significance of this trend with one or two objects, but when smaller objects are overwhelmingly found to be redshifted relative to their host galaxies, there can be only two explanations - either we are at most privileged position in the whole universe, where all smaller objects are preferentially running away from us heedless of the gravitational effects of their host galaxies OR these smaller objects have an additional intrinsic redshift that we don't yet understand. I prefer the latter.

 

[meteor]

"A catalogue of southern peculiar galaxies and associations" (Arp, 1987)
http://nedwww.ipac.caltech.edu/level5/SPGA_Atlas/frames.html

This catalogue includes galaxies that are apparently connected. Can be an useful source of information

Also, in post 119 of this thread
https://www.physicsforums.com/showthread.php?t=3593&page=3&pp=40
I give my explanation of the fact that two objects with different redshift seems to be connected

 

[Chronos]

How can you not give an 8% chance that anyone of 12 objects in a group is the least red shifted among the group? The main point of the Newman-Terzian paper was point out the flawed initial assumption Arp used. Intrinsic red shift is a very unsatisfactory explanation. Why would less massive bodies have more intrinsic red shift than the most massive body in a group? The preponderance of evidence supports the conclusion that z=distance. Arp's 388 potential 'exceptions' should not be given more credence than the overwhelming number of confirmations in support of expansion and the BB model. Exceptions, however, are interesting, and point the way to more refined explanations. That only means the current model is incomplete, not fundamentally wrong. Give credit where credit is due. We already know the current model has problems. Quantum physics and GR just don't mix without a quantum theory of gravity. I don't like the idea of appealing to higher orders of reality to explain the problem. Adding strings and branes, imo, only creates more difficulties than they resolve... 'mommy, where did strings come from?'. I think it is a mistake to stray from the observable universe into the unobservable to find answers. It flys into the face of science as I know it.

 

[turbo]

How can you not give an 8% chance that anyone of 12 objects in a group is the least red shifted among the group? The main point of the Newman-Terzian paper was point out the flawed initial assumption Arp used.

The problem with the "statistical treatment" of the example is that the authors of the paper stripped out causality with regard to the gravitational influence of the largest body - it was ignored completely. If you were to make some observations relating to the movements of bodies in our solar system, surely you would not use a method that disregards the gravitational influence of the sun! This is exactly what the authors of that paper did.

To restate:

There are eleven relatively small objects in the domain of one very large massive object.

The motions of the small objects orbiting the massive host object have a 50:50 chance of being toward us or away from us relative to the host.

If the host has one object it around it, there is one chance in two (1/2) that the object would be redshifted with respect to the host. If there are two objects, there is one chance in four (1/4) that both objects would be redshifted relative to the host. To save time, we will extrapolate: if there are eleven objects around the host, there is only one chance in 4096 that all eleven objects will be found to be redshifted relative to the host.

This is very basic statement of probability. If there is 1/x probabilty of an object being in a particular state, the probability of a set of objects with y members all being in that state at one time is 1/(x to the y power). It's hard for me to understand how the Newman-Terzian paper was published with such a very glaring basic statistical error.

Intrinsic red shift is a very unsatisfactory explanation.

The explanation may be unsatisfactory, but that does not invalidate the observations. The scientific method requires that if we cannot disprove the observations, we must come up with a model that explains them, and use the model to make testable predictions that can be used to verify, disprove, or refine the model.

 

[Chronos]

Turbo: Objections noted. Not all apparent redshift anomalies have been satisfactoriy explained using the standard model. Progress is, however, being made. So far as Newman-Terzian is concerned, I am not aware of any published studies refuting their position. I would be interested if you know of any references.

 

[Chronos]

Hubble, Galaxies Across Time and Space is an IMAX film based on the GOODS survey.

As there are ~30,000 galaxy images in GOODS, of which ~11,000 have measured redshifts, this might be a good test of the Arp hypothesis - how many 'discordant redshifts' appear in the GOODS data? AFAIK, all results are in the public domain, so anyone can check the selection methods used, for determining what constitutes a 'galaxy', and for interpreting spectra in terms of redshift.

In some ways it's a pity this is already 'finished'; otherwise we could ask Arp (or supporters) what they would expect to find in such a survey, and compare their predictions with the observations.

You may find this interesting: http://arxiv.org/PS_cache/astro-ph/pdf/0208/0208117.pdf

 

[turbo]

Thanks for posting that paper. Similar to your discomfort with intrinsic redshifts, I have a problem getting comfortable with the concept of quantized redshifts. My gut feeling is that there is a very simple understandable reason for discrepant redshifts, and that once we figure out the mechanics, the cause will be something that causes a smooth continuum of redshift, not a stepwise progression.

 

[Nereid]

To restate:

There are eleven relatively small objects in the domain of one very large massive object.

The motions of the small objects orbiting the massive host object have a 50:50 chance of being toward us or away from us relative to the host.

If the host has one object it around it, there is one chance in two (1/2) that the object would be redshifted with respect to the host. If there are two objects, there is one chance in four (1/4) that both objects would be redshifted relative to the host. To save time, we will extrapolate: if there are eleven objects around the host, there is only one chance in 4096 that all eleven objects will be found to be redshifted relative to the host.

This is very basic statement of probability. If there is 1/x probabilty of an object being in a particular state, the probability of a set of objects with y members all being in that state at one time is 1/(x to the y power). It's hard for me to understand how the Newman-Terzian paper was published with such a very glaring basic statistical error.

I'll get back to this in more detail when I've read more, but just quickly ... the tidal locking of Io, Europa, and Ganymede means that the probabilities of 'N moving towards' vs 'M moving away' is not random, so the probabilities turbo-1 describes would not apply (actually, *could* not apply). There's also measurement error - how many 'small' objects would show essentially zero redshift (wrt something), within the measurement error?

 

[turbo]

Nereid, I'll willingly play the devil's advocate on this one. What if you have a cloud of smaller objects that are essentially captivated and falling directly into the larger object without significant orbital motion? Some objects will be redshifted, a smaller number will be blueshifted (because we will miss a portion of those objects due to obscuration by the host galaxy) and those objects that are dropping into the host galaxy on a plane perpendicular to our line of sight will exhibit the same redshift as the host.

Assuming NO orbital motion (direct infall only), and an obsuration of blueshifted objects of 10%, and non-discrepant redshift due to perpendicular orientation of infall paths of some objects (maybe another 10%), we still have a probability of less than one in 512 that all the small objects that surround the massive host will exhibit an excess redshift relative to the host. This is a far cry from 8% (1/12) cited in Nemnam-Terzian.

I believe that these are very fair, conservative estimates. I don't want to paint you into a corner and try to make you defend the Newman-Terzian paper, but what do you think of their paper?

 

[Chronos]

respect

Thanks for posting that paper. Similar to your discomfort with intrinsic redshifts, I have a problem getting comfortable with the concept of quantized redshifts. My gut feeling is that there is a very simple understandable reason for discrepant redshifts, and that once we figure out the mechanics, the cause will be something that causes a smooth continuum of redshift, not a stepwise progression.

we are in agreement. i respect your objections and agree that current models are lacking. we need better predictive models to sort it out. something is still missing. the universe is a maddening puzzle. it taunts and teases us, but, resists yielding its secrets. i do believe, however, current models are more correct than incorrect. the evidence appears, at least to me, to be overwhelming. the piece still missing, imo, is quantum gravity.

are there observations that contradict expansion as the 'big picture' of the universe? of course. information theory [shannon and godel] insist this will always be the case. will someone find a paradigm shift to further advance our understanding of how the universe works? i hope so. i object to being trapped within a place as small as our solar system. that is what current theory condemns us to and i am not comfortable with that either. if we can see it, we should be able to explore it.

 

[Nereid]

Mesolensing is in between microlensing (point-like mass) and macrolensing (galactic-like mass) and has been proposed as a means by which to indirectly observe CDM. The means proposed is to observe multipli-imaged quasars and to record any phasing anomalies, and from those infer the mass and/or extent of the CDM doing the lensing.

King objects aren't a class or type of object, if I understand correctly, they are objects that conform to a particular model of mass distribution, and they have been invoked in the form of galactic halos to explain away the statistical preference in some studies for quasars to appear in closer (angular) proximity to galaxies than can be accounted for by chance.

Am working my way slowly through the posts in this thread, and have got up to here.

Several questions, and possible lines of investigation, occur to me:
- examine the Arp hypothesis (-ses?), that AGN galaxies (inc Seyferts) eject quasars (yes, that's an oversimplification); that the preponderance of galaxy/quasar discordant redshifts also involve ULX. I'm not sure yet what to do about his galaxy/galaxy discordant redshift observations.
- take a closer look at gravitational lensing - micro, meso, and 'traditional'
- GOODS, 2dF, and SDSS as tests of Arp's ideas. e.g. if there are ~300 'Arp' objects among the ~23k de Vaucouleur catalogue (RC3), then there should be ~300 among the 30k GOODS galaxies, and ~3k among the ~250k 2dF ones, and ~10k among the ~1m Subaru/XMM-Newton ones (many assumptions behind these statements!); are there?
- both 2dF and SDSS should permit a far more rigorous testing of any quasar/(active) galaxy association than finding interesting objects by eye and taking a deeper look; medium-deep surveys should also provide good tests
- among objects whose distance has been established by means independent of redshift, there will likely be some from Arp's catalogue, and surely some Seyferts (etc). What about associated discordant redshift objects? E.g. an SN in both objects of such a pair? We can look for ourselves, because there's a pretty complete list of all SN available on the internet.

Finally, some clarifications on King objects, meso-lensing, CDM cosmology, etc.

One of the challenges for the concordance model ('$$\Lambda CDM$$') is that it predicts far more 'small(ish) galaxies' than are actually observed, IIRC between 1 and 2 OOM too many. Now the model doesn't say what sort of things these objects should be, just (approx) what their mass function should be (how many of mass M₁, how many of mass M₂, etc), and (approx) the distribution of mass within them (hence 'King objects'; OK some oversimplification here too). Most importantly, it doesn't distiguish between baryonic matter and dark matter! Enter theories of galaxy evolution, and where it starts to get messy; enter observations, and where it starts to get interesting! Even within the Local Group - the ~20 (30?) galaxies dominated by the Milky Way and M31 - we keep finding more and more things, such as dwarf galaxies like Andromeda 9, dwarf galaxies being shredded by the MW, and inter-galactic clouds of gas. But the critical thing is the DM, which, because it's dark, can't be 'seen' in X-rays, the optical, radio, etc. However, DM objects can be detected through gravitational lensing; the question is, what sort of gravitational lens signature would a DM object leave on the light (radio, X-rays, etc) from a distant quasar? Or, within an Arp hypothesis, what should the results of observations designed to find 'meso-lensing by King objects' be?

 

[turbo]

I wish I had more time... Suggestions, anyone?

There is such a ton of research material available...

Nereid, your suggestions re: SN in Arp objects led me to this great SN list

http://web.pd.astro.it/supern/snear.txt

and I started comparing it to the Arp catalog, which is a pain in the butt. Differences in nomenclature make it very likely that I would miss Arp objects in the SN catalog.

http://members.aol.com/arpgalaxy/arplist.html

I decided to cut to the chase and Googled supernova and Arp and found this article about multiple SN in Arp 299:

http://www.space.com/scienceastronomy/aas202_supernova_030527.html

Which led me to explore Arp 299 in more detail. There is wealth of material in NED, of course, and there is a nice Arp 299 page here:

http://www.cv.nrao.edu/~jhibbard/a299/a299.html

It seems that Arp 299 might be a little "too" complex for the SN data to be really useful, as evidenced by this really busy paper:

http://www.journals.uchicago.edu/ApJ/journal/issues/ApJ/v532n2/50399/50399.text.html

I will try to locate a more "normal" Arp pair with SN data available. Can anyone suggest a more productive approach to locating SN data in apparently-interacting objects with discordant redshifts?

 

[Nereid]

The list of SN that you found seems similar to the 'official' one, associated with the IAU circulars (CBAT - Central Bureau of Astronomical Telegrams - is currently administered by the Smithsonian Astrophysical Observatory?); http://cfa-www.harvard.edu/iau/lists/RecentSupernovae.html is their list of SN since the start of 2003 (there's also a link to all SN).

You may find the best way to find a match is to put the list into a database or spreadsheet, sort by RA (or dec), filter by close matches to Arp objects (I assume you have a list of their RAs and decs), and examine each such in detail.

Be aware that even the IAU list has errors in it! They seem to be mostly typos (e.g. transposed digits).

 

[Chronos]

well, i hoped to blow the whole arp thing out of the equation. the preponderance of evidence says arp is wrong. i referenced two papers that explain why i am persuaded to think how i do. if you wish to change my mind, offer me papers that refute them. i have great respect for terzian as a scientist. his credentials are impeccable. he has published many papers and is highly respected in the scientific community. no one, to my knowledge, has ever refuted his papers. you don't achieve the honorary chair of cornell university by being a quack. apologies. if mr terzian cared to speak here, we should listen.

 

[turbo]

well, i hoped to blow the whole arp thing out of the equation. the preponderance of evidence says arp is wrong. i referenced two papers that explain why i am persuaded to think how i do. if you wish to change my mind, offer me papers that refute them. i have great respect for terzian as a scientist. his credentials are impeccable. he has published many papers and is highly respected in the scientific community. no one, to my knowledge, has ever refuted his papers. you don't achieve the honorary chair of cornell university by being a quack. apologies. if mr terzian cared to speak here, we should listen.

You don't need someone with a Phd to refute the Newman-Terzian paper for you. Trust your own abilities. Read the paper and you will see that the authors (with lots of hand-waving and misdirection) ignored the central point of Arp's argument - that the smaller bodies are under the gravitational influence of the large body and therefore must exhibit some predictable behavior in their motions relative to the host. After butchering the model by ignoring gravity, they then treated the redshift relationships as a matter of simple ordination, saying that the bodies are all essentially equivalent, and that there is a 1 in 12 chance (8%) that the largest galaxy will have the smallest redshift. Those methods are so basically wrong, the paper should never have gotten past the referees.

Just sketch out the model on a piece of paper (large galaxy surrounded by 11 smaller bodies), assign some orbits to the small bodies and then imagine the coincidence (one in 4096) that would have to occur for ALL the smaller bodies to be moving away from us relative to the host. Here's a wrinkle: When you sketch out the model, assign one of the bodies an orbit that is about perpendicular to our line of sight. Now, try to arrange for that object to have a redshift relative to the host without invoking an intrinsic redshift of some kind.

Once you have sketched out the model and done some thought-experiments with it, re-read the Newman-Terzian paper. The flaws in their analysis will become immediately apparent. It's not rocket science.

Added 2:40pm... Let's do a really simplified example, with all objects in about the same plane around a central body - just like our own solar system. To an observer outside our "solar system" each "planet" would have measurable redshift relative to the sun less than 50% of the time. There would be no measurable redshift near transit or occultation and there would be blueshift on the other side of the orbit. If the outside observer noted ALL NINE "planets" being redshifted relative to the sun he would surmise (and rightly so) that he was observing a very special and rare alignment, with less than one chance in 1024 of ocurring at the particular time of his observation. He would not say "well, there are 10 objects in that system, so there is one chance in ten that the central object would have the least amount of redshift." At least he wouldn't say that if he had any sense. The Newman-Terzian paper uses exactly that approach to "refute" Arp. You don't need a Phd to invalidate the Newman-Terzian paper, just grade-school math and some common sense.

 

[Chronos]

Orientation could matter. If the axis of rotation is pointed at the observer, all bodies in the associated group would have indistinguishable redshifts. The lack of relative blue shifts is more puzzling than explanatory

 

[turbo]

Orientation could matter. If the axis of rotation is pointed at the observer, all bodies in the associated group would have indistinguishable redshifts. The lack of relative blue shifts is more puzzling than explanatory

Exactly! When orbit inclination is taken into account, we should expect the smaller objects to have significant redshifts with respect to the host less than half the time. Finding 11 out of 11 small objects around a large galaxy to be redshifted relative to the host is actually much more implausible than Arp calculated. He used a simple 50:50 (redshift-blueshift) calculation, which is generously conservative.

 

[shrumeo]

why were they assuming that the smaller bodies are orbiting the larger galaxy?
what was the indication that they were gravitationally associated?
I forgot, sorry.

 

[shrumeo]

Apologies for my ignorance!

Sorry for mixing threads but I hope someone could clear this up.

This came from "Seeing Red" thread:

In this page comes the formula for the calculation of gravitational redshift for a photon emitted from a star
http://en.wikipedia.org/wiki/Gravitational_redshift

$$z= \frac {G}{c^2}* \frac{M}{R}$$

M is the mass of the star and R its radius

I guess I never thought to try something as simple as pluggin into this eq. so I tried it for a typical quasar (or at least how it's typically described). So I replied with:

so a quasar is supposed to have a mass comparable to a galaxy
i will say a typical galaxy holds 10b stars,
let's just say our hypothetical quasar has 10b solar masses
IIRC, a typical quasar has a radius comparable to the solar system
I will just say 30 AU

mass of sun = 2.3 x 10^33 g (from the Wikepedia page)
10b x this = 2.3 x 10^43 g
radius of 30 AU ~ 4.5 x 10^14 cm

plugging these into the equation above, I get a gravitational redshift from a typical quasar to be z=3.78 !

Please tell me if my assumptions are wrong.
Thanks

 

[shrumeo]

Oh, wait, they are wrong. I was using the approximation for a small star.
When you use the real eq. you realize there is a limit to M/R, otherwise you are a black hole. Sorry for blemishing this thread.
:)

 

[turbo]

NGC 450 and UGC 806 - interacting galaxies.GC 4

Here is a very compelling example of interacting galaxies that appear (if redshift=distance) to be simply a chance projection. The authors evaluate the pair in light of morphology, redshift, rotational speed, etc, and conclude that the galaxies must be interacting despite their widely disparate redshifts. Go to the page below and get the scanned refereed PDF article.

http://adsabs.harvard.edu/cgi-bin/nph-bib_query?1994ApJ...432..135M

There are a number of very good photographs of this pair on the 'net. Here is a page containing some interesting pictures of interacting galaxy pairs, including this one.

http://www.astr.ua.edu/pairs2.html

 

[Nereid]

Here is a very compelling example of interacting galaxies that appear (if redshift=distance) to be simply a chance projection. The authors evaluate the pair in light of morphology, redshift, rotational speed, etc, and conclude that the galaxies must be interacting despite their widely disparate redshifts. Go to the page below and get the scanned refereed PDF article.

http://adsabs.harvard.edu/cgi-bin/nph-bib_query?1994ApJ...432..135M

There are a number of very good photographs of this pair on the 'net. Here is a page containing some interesting pictures of interacting galaxy pairs, including this one.

http://www.astr.ua.edu/pairs2.html

Don't you mean UGC 807?

I note that the Arp et al paper which you cite was published in 1994; I also note that NGC 450 and UGC 807 were the subject of several Hubble observations - images and spectra - how do Arp et al's claims stack up in the light of the much higher resolution HST observations?

If anyone's intetested, http://archive.stsci.edu/cgi-bin/genlinks_search.cgi?target=NGC450&resolver=SIMBAD is a place you can start your search for raw data, to perform your own analyses ...

 

[turbo]

You're right of course. I slipped a number key (l-r) when typing - I'm an acceptable touch-typist most times, but have trouble with numbers, and didn't catch it in proofreading.

At any rate, UGC 807 (with the higher redshift) appears more likely to be in front of NGC 450 than in back of it, and there are morphological problems between the pair that point to interaction - these are not a chance projection.

http://nedwww.ipac.caltech.edu/level5/index.html

There are some really wonderful atlases here, where lots more of these conundrums can be found.

 

[turbo]

I note that the Arp et al paper which you cite was published in 1994; I also note that NGC 450 and UGC 807 were the subject of several Hubble observations - images and spectra - how do Arp et al's claims stack up in the light of the much higher resolution HST observations?

The exposures with HST's wide-field planetary camera are very crisp, but not deep. They don't respond well to my FITS viewer's contrast function, either. I would be very interested in seeing if Arp et al's redshift measurements of NGC 450's HI regions are confirmed, but don't know if this work has been done, or can be extracted from raw data already collected.

On another front:

Here is a link to a paper in which photometric measurements were done to determine the level at which NGC 450's disc diminishes the light of UGC 807. Interestingly they found no measurable extinction - even though there must be a 10x more massive halo of dark matter around NGC 450. I guess it MUST be non-baryonic matter.

http://adsbit.harvard.edu/cgi-bin/n...plate_select=NO&data_type=GIF&type=SCREEN_GIF

If indeed UGC 807 is slightly in front of NGC 450 (as it appears to me in deep exposures),

http://www.sdss.org/news/releases/20010605.edr.img18.html
http://160.114.99.91/astrojan/Ngc/ngc0450.gif


their result (no dimming) is to be expected, although it throws redshift=distance in the trash. It would be intersting to see a similar analyses to see if the light of NGC 450 suffers extinction due to absorption by the disc of UGC 807.

 

[Nereid]

Just for fun ...

Today's APOD is a very nice piccie of M57 (the Ring Nebula) and IC 1296. If the two had been just a few ' closer together - on the sky - would someone have considered it strong evidence of interaction?

For sure, no one (seriously) today, but in the first couple of decades of last century ...

Of course, Arp et al are careful to try to distinguish mere coincidences (not) like this from possibly real things (yes, I will continue reading the posts in this thread); beyond Arp, how does one go about clearly demonstrating 'mere' coincidence?

Let's also mention Stephan's Quintet. IIRC, turbo-1's earlier post in this thread on SQ implies that Arp (or colleagues, or turbo-1) insisting that NGC 7320 (the blue spiral) is actually interating with the other galaxies in the quintet. If this stance is still being taken, I guess we can conclude that one of the redshift-independent methods to determine distance is contra-indicated (according to the discordant redshift brigade), namely surface-brightness fluctuations; in the HST image, NGC 7320 clearly has greater such fluctuations than NGC 7318A, 7318B, and 7319. So we could reasonably ask, which of the generally accepted, non-redshift methods of estimating distances* DO they accept as valid?

*from the Freedman paper on the Hubble constant:
- Cepheids
- Type Ia supernovae
- Tully-Fisher relation
- surface-brightness fluctations
- Type II supernovae
- fundamental plane
- gravitations lensing
- Sunyaev-Zel'dovich.

 

[turbo]

Today's APOD is a very nice piccie of M57 (the Ring Nebula) and IC 1296. If the two had been just a few ' closer together - on the sky - would someone have considered it strong evidence of interaction?

Even if we had no clue about the nature of those two objects, there is nothing in their morphology (in this very beautiful image!) that suggests interaction. Anomalous star formation, tidal disruption, etc.

beyond Arp, how does one go about clearly demonstrating 'mere' coincidence?

There's the rub. The universe is a pretty big place so when a researcher points to an apparently-interacting pair of objects with discordant redshifts, conventional cosmologists can (very logically) say "given the number of objects in the universe, we should expect to see coincidental alignments like this" as they casually dismiss the observation. It is not incumbent on conventional cosmologists to demonstrate coincidence, nor is it politically expedient to spend much effort or time at it. When someone does take the time to critique the work of Arp, Burbidge, et al, their critiques are often accepted as fact, even when their analyses are badly flawed, like the Newman-Terzian paper cited in a post above.
http://adsabs.harvard.edu/cgi-bin/n...J...441..505N&db_key=AST&high=3325b47acc06425

Let's also mention Stephan's Quintet. IIRC, turbo-1's earlier post in this thread on SQ implies that Arp (or colleagues, or turbo-1) insisting that NGC 7320 (the blue spiral) is actually interating with the other galaxies in the quintet.

I didn't cite Stephan's Quintet as an example of interaction between discordant-redshift galaxies. I think the high resolution imagery from HST put that one to bed years ago. There are, however, examples of apparently-interacting objects with very discordant redshifts, one of which I mentioned in one of the first posts in this thread.

Seyfert galaxy NGC 7603 z=0.029 is connected to its apparently ejected companion z=0.057 by a luminous bridge in which are embedded two compact emission line objects of z=0.243 and 0.391. Scroll down to the images for a look, then read the paper.

http://arxiv.org/abs/astro-ph/0203466

Conservative astronomers claim that these compact high redshift objects are at the distances implied by the Hubble redshift model and the filament is coincidentally projected over them. They also claim that the companion 7603B is at the cosmological distance implied by its redshift which raises the question: Why is there a luminous bridge connecting the two objects if they are too far apart to interact?

It would be helpful for the reader to load the paper above and scroll down to the false-color images. Look at the morphology of the system and for a moment, pretend that the redshifts of these objects have not yet been measured. There is a luminous filament extending from the Seyfert to the ejected object. Most reasonable people would say that these objects are interacting, similar to the M51 system. Now, measure the redshifts of the objects. Suddenly, this obvious example of interaction/ejection turns out to be a chance projection with NO chance of interaction.

How can this be? It's because conventional cosmologists "know" with 100% certainty that the redshifts of these objects are due to cosmological expansion, that's why! There are no questions left unanswered about redshift. Well, there's the little question about why young supermassive stars in our own galaxy and others are streaming directly away from us in every direction (K-effect), but that's a minor quibble! We already know everything there is to know about redshift.

Sarcasm aside, if young supermassive stars in our own galaxy can exhibit intrinsic excess redshift, is it possible that some extra-galactic bodies might also have intrinsic redshifts? The orthodox astronomical community rejects that possibility out-of-hand, and that's bad science. It might be that Lopez-Corredoira and Gutierrez are wrong about NGC 7603. It might be that Arp et al are wrong about NGC 4319 and Markarian 205 and every other example of apparently-interacting objects with discordant redshifts. If only one of these numerous apparent interactions is real, however, (not ALL of them, only ONE of them) the simplistic redshift=distance rule is in for some serious modification, and we have a LOT of work to do.

 

[Nereid]

You may find this interesting: http://arxiv.org/PS_cache/astro-ph/pdf/0208/0208117.pdf

Thank you Chronos, a most interesting read.

This paper should allow us to put several 'association' and 'quantized z' hypotheses to sleep. Does anyone know if any serious astronomers in either of those camps are still banging those drums?

Now that some SDSS results are in the public domain, perhaps a similar piece of research could be done using those? The good news is that SDSS uses a completely different method to select objects for specta than 2dF did, so if similar analyses of the two datasets yield similar results there'll surely be no place to hide!

I particularly liked the approach taken here: a proponent in one camp suggests a method of analysis, and a neutral third party carried out the work, using publicly available data. What's good? The method and expected outcomes were clearly defined BEFORE the work was done, and the datasets are in the public domain (you don't like the conclusions? there's nothing at all stopping you from performing your own analyses!)

 

[Nereid]

Thank you for your time to consider these thought provoking questions.

Good questions, big egg. Some quick, but not necessarily complete, answers:

Are the disparate red shifts associated only with quasars and galaxies?

Are there coupling galaxies with disparate red shifts?

Arp et al (and turbo-1) would claim there are plenty of galaxy-galaxy ones; almost everyone else would say that any such are chance alignments (billions of galaxies detectable by the likes of the VLTs, Geminis, Kecks, HST, etc; chance alignments will occur by the thousands, maybe millions)

Are there supernovas in the proximity of the quasars?

turbo-1 is checking that; by simple random probability, there will surely be some. Of course, it also depends on what you mean by 'proximity'! A dedicated observational campaign would surely find lots of SN associated with quasars. http://snap.lbl.gov/ is one project that would help answer this; let's all work together to help ensure it gets off the ground!

What about the quantum red shifts? They will also create a hard problem to the BBT.

When data was sparse, you could crunch the numbers and show tantilizing hints of such quantization; with 2dF (and now SDSS) in hand, all such hints have evaporated.

Are there statistical steadies on disparate red shifts?

Yes; see the paper which Chronos provided a link to.

 

[Nereid]

http://www.naoj.org/Science/SubaruProject/SDS/scijust_qsos.html

There is ongoing work that can help eliminate the quasar-candidate selection bias between radio-active and radio-quiet objects. This research should help clear statistical questions surrounding the optical selection criteria for radio-quiet candidates.

Not only radio-loud and -quiet, but also (especially) optical biases. This work will also improve our confidence in the AGN model, per my earlier post.

 

[Nereid]

Nereid, I'll willingly play the devil's advocate on this one. What if you have a cloud of smaller objects that are essentially captivated and falling directly into the larger object without significant orbital motion? Some objects will be redshifted, a smaller number will be blueshifted (because we will miss a portion of those objects due to obscuration by the host galaxy) and those objects that are dropping into the host galaxy on a plane perpendicular to our line of sight will exhibit the same redshift as the host.

Assuming NO orbital motion (direct infall only), and an obsuration of blueshifted objects of 10%, and non-discrepant redshift due to perpendicular orientation of infall paths of some objects (maybe another 10%), we still have a probability of less than one in 512 that all the small objects that surround the massive host will exhibit an excess redshift relative to the host. This is a far cry from 8% (1/12) cited in Nemnam-Terzian.

I believe that these are very fair, conservative estimates. I don't want to paint you into a corner and try to make you defend the Newman-Terzian paper, but what do you think of their paper?

I think the paper is excellent. (I haven't been able to find an on-line version of the original Arp paper, so my comments must be rather tentative).

The main gulf seems to be the underlying assumptions; Newman & Terzian (N&T) make none whatsoever about the dynamical relationship among the Local Group objects; you (and presumably Arp) assume a single, gravitationally domanating object (M31).

N&T actually highlight this gulf, before doing their own analysis: "The simple dynamical picture presented by Arp is inconsistent with the observed disposition of the galaxies of the Local Group. There is no dominant galaxy per se, and the dynamical picture presented by the group's members is highly complex - indeed, the potential for chaotic behavior in this dynamical system cannot be excluded."

It's hard for me to understand how the Newman-Terzian paper was published with such a very glaring basic statistical error.

My guess is that, as they stated their assumptions clearly, there was no 'glaring basic stastical error'; perhaps we could move the debate to the question of the likely dynamical status of the Local Group?

Since N&T, new LG members have been identified, and (maybe) better redshifts obtained for those members. Using http://www.seds.org/messier/more/local.html , we find the following members of the M31 sub-group: M32, M110, NGC 147, NGC 185, And I, And II, And III, And IV, Pegasus dwarf, Cassiopia dwarf, and And VIII (And IV maybe?; this site was last updated in 2003, before And9 was discovered). Here are the stated relative velocities (cf MW centre, in km/s) of those (not all M31 sub-group members have published redshifts, apparently):
M31: -59
M32: +35
M110: -1
NGC 147: +89
NGC 185: +39
And VIII: -250

So, 4 have a +ve RV cf M31's, and 1 a -ve RV. Even if this sub-group could be shown to be quite isolated (it's not at all; M33 and the Milky Way must have a significant influence, without even looking at any possible DM concentrations, or high velocity gas clouds), 1/5 isn't all that improbable.

(Next, MW sub-group; if I have time).

 

[turbo]

Hi Nereid!

As you know, when I posted my critique of the Newman-Terzian paper, I did not dispute the observations quoted by either side. I kept my post focused on the statistical methods employed by both sides. The problem with the Newman-Terzian paper is that the authors stripped the gravitational influence of the largest body out of the analysis (which was a central point in Arp's presentation) and then presented the 12-body problem as a simple case of ordination. This is the best way to understate the odds that all eleven small companions would be reshifted relative to the largest body. The astronomical community gave this paper free pass, although any sharp 7th grader would have trouble letting it by.

In essence, the authors presented Arp's system of 12 bodies as a simple set of equivalent items, ignoring the gravitational influence of the (FAR more) massive body then said "there is a 1/12 chance that the largest body will be the least redshifted". This is VERY wrong. It is like a pair of cops barging into a room looking for a bank robber and finding eleven pre-schoolers and one adult with stain from a dye-bomb on him and concluding "there is a one in twelve chance that this is the person who robbed the bank".

As you point out, the authors state that "the dynamical picture presented by the group's members is highly complex - indeed, the potential for chaotic behavior in this dynamical system cannot be excluded". That statement does not excuse them from ignoring basic probability, however. In a chaotic environment, should we expect to find ALL the smaller bodies to be redshifted with respect to the largest body "some of the time", "most of the time", or "all of the time"? Chaotic behavior in complex systems permits short-term anomolies (even very unlikely ones) but it does not trump probability in the long term or over multiple samples.

To summarize again, N-T did not dispute Arp's redshift characterizations of either of the galaxy groups (2 groups, each with eleven companions around more massive galaxies) nor did I. They challenged Arp's statistical methodology, and they presented alternative "analyses" that were very badly flawed, even cynical. Arp-bashing is popular and easy, but I think you would have trouble getting the authors to defend that paper today.

 

[Chronos]

Nereid, I was beginning to think I was alone in my interpretation of the NT paper. I believe it is very relevant and entirely valid. I have not heard anyone claim, and have no reason to dispute, that Mr Arp is other than a well intentioned and honorable man. I just think he is wrong and can't admit it. My objection to Arp et al is in using mass media in a transparently lame attempt to convince the public that flawed assumptions and flawed math are attributable to a great conspiracy by 'mainstream scientists' to suppress their 'discoveries'. True science is hard enough without people appealing to public opinion to support their failed theories. When peer opinion, and observational evidence is overwhelmingly against you, the honorable thing to do is admit it. Einstein did. As it turns out, he may not have been the least bit wrong. His concept of a 'cosmological constant' is very much alive these days. But he had enough humility and respect for his peers to admit he did not have the math to prove it.

 

[Chronos]

Turbo, please elaborate. In what way is Arp's statistical analysis more 'correct' than NT's? NT made no assumptions about system dynamics, Arp did. Arp offered no proof of this basic assumption. It is therefore suspect, at best, and mostly contraindicated by observational evidence. I am not a dogmatic fool, but, I am a sucker for observational facts and supporting math. So, from my reference frame, the Arp assumption is invalid.