More from Arp et al.

[Thanatos]

I am still stuck on this [post 2628]

. . . I didn't calculate a single redshift distance for Virgo Cluster Galaxies. The paper utilizes the Tully-Fisher Relation. The Tully-Fisher relation is a tight relationship between the absolute magnitude and the rotational velocity of a galaxy. No redshift distances are needed. So the negative redshifts are irrelevant to the distance calculation. The redshifts are used to calculate the PV72 value:

PV72 = Vvir - DistancexH0

What do you mean by 'no redshift distances are needed' to calculate the PV72 value'? The 'V' in Vvir is, according to LEDA:

" . . . The heliocentric cz, v is computed from the data in the compilation of redshifts by rejecting outliers and measurements marked for their bad quality and computing a weighted average of the remaining measurements.
The heliocentric cz, v is used to compute the redshift in other reference frames: Namely in the Local Group referential, vlg, the Galactic referential, vgsr, the Virgo cluster referential, vvir, and the CMB referential, v3k."

[dgruss23]

I am still stuck on this [post 2628]
What do you mean by 'no redshift distances are needed' to calculate the PV72 value'? The 'V' in Vvir is, according to LEDA:

" . . . The heliocentric cz, v is computed from the data in the compilation of redshifts by rejecting outliers and measurements marked for their bad quality and computing a weighted average of the remaining measurements.
The heliocentric cz, v is used to compute the redshift in other reference frames: Namely in the Local Group referential, vlg, the Galactic referential, vgsr, the Virgo cluster referential, vvir, and the CMB referential, v3k."

There is a difference between a "redshift" and a "redshift distance". The redshift is specifically a shift of the observed spectral lines to longer wavelengths. The relative shift is typically expressed as "z" where:

 z = Wavelength observed - wavelength actual
                           Wavelength actual

This is the direct measurement of the amount of spectral shift. It is a redshift when the observed wavelength is longer than the actual wavelength.

Under the assumption that the observed wavelength shifts may be treated as velocities the spectral shift (z) may be multiplied by the speed of light (c) to get a velocity redshift:

Velocity = cz

In the conversation we're having here redshift may mean the spectral shift (z) or it may mean the velocity (cz). The "redshift" is not a distance.

In order to derive a "redshift distance" you must have a "redshift velocity" and a value for the Hubble constant (H0).

Then the redshift distance becomes:

Redshift Distance = cz/H0

But as I've previously explained, you don't need to calculate redshift distance to get a value for PV72. If you have a redshift independent distance (such as Cepheid's, TFR ...) you can still get a PV72 value by the following:

PV72 = Vvir - (Distance x H0)

where distance in the above equation may be derived from Cepheids, the TFR or any of the other secondary distance indicators.

For example, NGC 4535 has a Cepheid distance of 15.8 Mpc and a redshift velocity corrected to the Virgocentric reference frame of 2029 km s-1.

So plugging those numbers and a Hubble constant of 72 gives PV72 = +891 km s-1.

[Thanatos]

Would you agree that redshift is a measured value and redshift velocity is a derivation based on that measurement? I have no problem considering there may be discrepancies between the TF distance and redshift distances, just your resistance to making comparisons. I assumed you were asserting a correlation between morphology and redshift. I randomly selected 100 galaxies that met your published selection criteria and found no statistically signficant correlations.

[dgruss23]

Would you agree that redshift is a measured value and redshift velocity is a derivation based on that measurement?

Of course. Mainstream astronomers use redshift velocity all the time. Are you suggesting multiplying the spectral shift by "c" introduces a problem? If so, then that's a mainstream problem too.

I have no problem considering there may be discrepancies between the TF distance and redshift distances, just your resistance to making comparisons.

Thanatos, you're illustrating again that you don't know what you're talking about. I have no "resistance" to calculating redshift distances. There simply is no need to calculate redshift distances because you get the exact same numbers if you calculate PV72 the way I did, that you would get if you first calculate a redshift distance a TF distance and then calculate the redshift discrepancy from the difference between the two.

If you want to calculate redshift distances, then this is how you will calculate the redshift discrepancy.

Redshift velocity difference = (Hubble distance - TFR distance) x H0

For NGC 4535 used in my last post the Vvir redshift velocity is 2029 km s-1. Divide this by a Hubble constant of 72 km s-1 Mpc-1 and you get a redshift distance of 28.18 Mpc. So then the redshift velocity difference (RVD) is:

RVD = (28.18 Mpc - 15.8 Mpc) x 72 km s-1 Mpc-1 = 891 km s-1

That may be compared with the PV72 value of 891 km s-1 calculated in the previous post.

Do you understand now?

I assumed you were asserting a correlation between morphology and redshift. I randomly selected 100 galaxies that met your published selection criteria and found no statistically signficant correlations.

I already responded to this in post #2628 . Your approach is not a correct procedure. You must calculate redshift independent distance to make such a comparison ... as I said:

You can't do it that way Thanatos. That's an incorrect procedure and I wouldn't expect you to find any correlation between redshift and morphology by that means. You need actual distances to galaxies calculated by redshift independent means (such as the TFR) or you can go without distances if you have galaxies in the same cluster ... but you're always in better shape if you apply the TFR.

[Thanatos]

I appreciate you taking the time to point out how I missed the boat with my naive questions.

[dgruss23]

I appreciate you taking the time to point out how I missed the boat with my naive questions.

Not "naive questions", but certainly too not questions asked without a bit of infused accusation. You say I'm resistant to calculating redshift distances when you never asked "Why don't you calculate redshift distances?" If that is what you wanted to know, just ask it instead of accusing me of resisting it, cherry picking ...

[Nereid]

TomT's (#2662):

So before going further, I have 2 questions:

(1) dgruss23 is the above list the same as used for Table II, save for the 1 galaxy I missed?

(2) Thanatos, do you have any galaxies you feel should be added to this list? If so, could you state which ones, and why they should be included.

I've looked, but can't find any posts in which these two questions were answered.

Nereid's (#2678):

So in order to check the numbers in your post, one would need the names of (another) 56 galaxies (18+15+23), is that correct?

Ditto, was this question answered?

[Nereid]

The paper's selection criteria are somewhat vague (my bold):What was the (unstated?) selection criterion (criteria?) you used, wrt "in the Virgo cluster" vs that used by Thanatos? Specifically, what query (or queries) did you submit to LEDA?

Sandage et al provided the list of spiral galaxies from the complete Virgo cluster sample. I used their list and searched the LEDA database for the galaxies that met the criteria stated in the paper as the starting point. Since Sandage et al were working from the complete VCC and since I'm familiar with the other papers that have applied the TFR to Virgo, I didn't see any need to SQL HyperLeda.

However, you seem capable of scouring that VCC list for an example that was missed. You can then plug that in with the rest of the sample to see if that changes the numbers. I've already done that and summarized the results in several posts in response to you and Thanatos. See here .

Thanks.

In light of the difficulties you have had with LEDA, (as you say) the wavelength-dependence of morphological classification, the well-known inhomogeneous structure of the Virgo cluster, the sensitivity of your "intrinsic redshift" conclusions to re-classifications (wrt the Virgo cluster membership, morphology and luminosity type, and just plain small number statistics), and so on, it may be a good idea to hope that DIVES (Deep Infrared Virgo Environments Survey) gets approved (the link is to a >1MB PDF file).

[dgruss23]

Nereid's (#2678)Ditto, was this question answered?

There were 49 spirals in the Virgo cluster sample - Table II of the paper. I believe the 18 you're referring to is part of the Elliptical galaxy sample. So the answer to your questions is "No", you do not need the identifications of "another" 56 galaxies.

[dgruss23]

Thanks.

In light of the difficulties you have had with LEDA, (as you say) the wavelength-dependence of morphological classification, the well-known inhomogeneous structure of the Virgo cluster, the sensitivity of your "intrinsic redshift" conclusions to re-classifications (wrt the Virgo cluster membership, morphology and luminosity type, and just plain small number statistics), and so on, it may be a good idea to hope that DIVES (Deep Infrared Virgo Environments Survey) gets approved (the link is to a >1MB PDF file).

There are a few misconceptions here:

1. Wavelength-dependence of morphological classification. Where the TFR is concerned, the classification in optical wavelenths leads to the same breakdown in both the B-band as the K-band. So this is not an issue.

2. As I've already detailed in a number of posts ( here and here and here , the ~ 2000 galaxies ends up being reduced to the relatively small final sample of spirals utilized in the paper for a variety of legitimate reasons. The DIVES is not going to change the Virgo results because there is not a significant population of ScI galaxies to add to the sample. Those that might theoretically be added would only re-inforce the systematic excess redshift for ScI's that is discussed.

3. Small number statistics. I can't help that Virgo doesn't have more ScI's. The stystematic excess redshift for the ScI's in Virgo is unexpected because they have a distance distribution and coordinate distribution consistent with the ellipticals in the cluster. It simply won't work to suggest they're all in one filament infalling together.

And as already noted there are other parts to the series of papers that utilize much larger samples. So far none of that has been discussed.

As for changes in the LEDA database, through all those changes the Virgo results remain unchanged. Sure this or that galaxy may be added or taken out of the sample because of the selection criteria, but the systematic excess redshift of the ScI's is always there.

[Thanatos]

Lets talk sensibly, not argue. I wanted to understand your paper, not argue like this. Perhaps this is not the right forum to beg to differ.

[VanderL]

I may have missed it, but it seems that after Dgruss23 explained all the details, there is nothing in Nereid's selection bias investigation (or any other challenger) that would have a strong impact on the conclusions of the Russell paper. As far as I can see there's not a single reason not to accept Russell's conclusions...

Yes, you missed the numerous reasons to be skeptical of Russell's conclusions.

After seeing the last pages of exchanges, the answer is no, I didn't miss any reasons.

I think there is every reason to accept the possibility of intrinsic redshift based on this paper.

"Possibility" is one thing. Actuality is quite another.

While "possibility" is a good result for an ATM concept, I think that
based on the limits of current available data (which as Dgruss23 says will not be likely to change), and based on the failure of challengers to argue any dents in Russell's analysis, we'll have to change "possibility" to "reality".

Cheers.

[Nereid]

Nereid's (#2678)Ditto, was this question answered?

There were 49 spirals in the Virgo cluster sample - Table II of the paper. I believe the 18 you're referring to is part of the Elliptical galaxy sample. So the answer to your questions is "No", you do not need the identifications of "another" 56 galaxies.

Here is the original question (post #2675):

[snip]

Without calculating TFR distances, let's just assume they are cluster members and calculate the mean redshift breaking them down by morphology:

There were 4 ScI's/Seyferts among these with a mean redshift of 2276 km s-1 (Table II group mean was 1855 km s-1).

There were 10 Sbc/ScIII's with a mean redshift of 1537 km s-1 (Table II group mean was 1268 km s-1)

There were 4 Sab/Sb's with a mean redshift of 1029 km s-1 (Table II group mean was 913 km s-1).

[snip]

I'm having difficulty with these numbers, can you help please?

By referring to Table II, I deduce that "redshift" means Vvir - is that correct?

The 11 galaxies which comprise "ScI's/Seyferts" in Table II (on the line labelled "Sbc and Sc I/I-II and Seyferts") are: "6 ScI's, 3 Sab/Sb Seyferts, and 2 ScII's: NGC 4254, NGC 4303, NGC 4321, NGC 4501, NGC 4535, and NGC 4536 are the ScI's; NGc 4388, NGC 4450, and NGC 4579 are the Sab/Sb Seyferts; NGC 4651 and NGC 4654 are the ScII's" (source); is that correct?

There are 15 "Sab/Sb's" in Table II (on the line labelled "SO-a to Sb") - these are not named in the paper, nor have they been listed in any post in this thread; is that correct?

There are 23 "Sbc/ScIII's" in Table II (on the line labelled "Sbc/ScII-III to IV") - these are not named in the paper, nor have they been listed in any post in this thread; is that correct?

Of the 11 "ScI's/Seyferts", 9 are in Table III, one is among Thanatos' "list of 92" (NGC 4654) and one is not in either list (NGC 4651); is that correct?

You answered "Yes" to each question.

The 15 "Sab/Sb's" in Table II and the 23 "Sbc/ScIII's" in Table II have not been named (so far).

Where does the "18" come from? dgrus23's post #2632 (my bold):

Finally there were 18 galaxies with inclinations that did not meet the criteria because they were too highly inclined or to close to face on.

[snip]

There were 4 ScI's/Seyferts among these with a mean redshift of 2276 km s-1 (Table II group mean was 1855 km s-1).

There were 10 Sbc/ScIII's with a mean redshift of 1537 km s-1 (Table II group mean was 1268 km s-1)

There were 4 Sab/Sb's with a mean redshift of 1029 km s-1 (Table II group mean was 913 km s-1).

Could you please provide a list of the names of these 56 galaxies?

[Nereid]

[snip]

As for changes in the LEDA database, through all those changes the Virgo results remain unchanged. Sure this or that galaxy may be added or taken out of the sample because of the selection criteria, but the systematic excess redshift of the ScI's is always there.

Just quickly ... one (of several) shortcomings in that part of the paper I have been examining is the lack of any error analysis, for the results presented in Tables II and III. Without such an analysis, the conclusions must be more tentative than would be case if there had been such an analysis.

In particular, the absence of such an analysis prevents any meaningful estimates of the statistical likelihood of the conclusions ... under almost any set of assumptions regarding the galaxies selected (or not).

[dgruss23]

Here is the original question (post #2675):You answered "Yes" to each question.

The 15 "Sab/Sb's" in Table II and the 23 "Sbc/ScIII's" in Table II have not been named (so far).

Where does the "18" come from? dgrus23's post #2632 (my bold):Could you please provide a list of the names of these 56 galaxies?

The problem here is that there are two different sets of selection criteria at work. There is the criteria I used in the paper, and then there is the criteria I would recommend today based upon additional studies (mine and other researchers). None of those revisions to selection criteria would be based upon the selection of Virgo galaxies specifically - just based upon what research shows about developing a TFR sample with the smallest possible uncertainty. So some of the "18" were in the list of used in the paper because they were not excluded at that time. There is also the problem of the LEDA database updates we've discussed.

So, I have no problem providing you with the list, but I think you need to go through the Virgo list yourself and come to your own conclusion. This is important. I'm not going through the list again. I already did that a month ago (post #2632 and 2642) and got limited response at that time. You are the one arguing that there may be other examples that will alter the redshift discrepancy within Virgo. So you do the work this time. I already tried to illustrate to you with the Virgo list of ~2000 Thanatos linked to and got minimal response and the responses I did get/am getting certainly fall outside the ATM guidelines for responding in a timely fashion as illustrated by the fact that a month after I did that work we are now re-hashing it.

I will of course reply to what you come up with and give you advice/assistance/clarification if you need it. You can follow my posts from a month+ ago if you want to understand what I consider the proper selection criteria.

So here is the list of 49 spiral galaxies from the paper:

11 ScI's/Seyferts:

NGC 4654, N4651, N4303, N4388, N4450, N4501, N4579, N4254, N4321, N4535, N4536

23 Sbc/Sc:

N4420, N4689, N4123, N4701, U7943, N4356, N4522, N4207, N4390, N4294, N4647, IC3742, N4189, N4212, N4298, U7676, N4540, N4498, N4206, N4237, N4451, N4639, N4527

15 SO-a/Sab/Sb

N4192, N4216, N4343, N4380, N4402, N4569, N4698, N4457, N4580, N4606, IC3392, N4383, N4293, N4336, N4548

Understand that the "18" includes a mixture of galaxies that were originally rejected and that would be rejected according the current LEDA updates and the newer selection criteria I provided on the list. This list is the 18 that would be rejected based upon inclination according to the newere criteria and therefore some of the following are duplicates with the above:

18 rejected based upon inclination (based upon the ~2000 Virgo galaxies list Thanatos linked to):

IC 3061, N4207, N4235, N4206, N4276, N4289, N4302, N4303, N4307, N4313, N4316, IC3322a, N4394, N4412, N4416, N4430, N4445, IC3727

[dgruss23]

Just quickly ... one (of several) shortcomings in that part of the paper I have been examining is the lack of any error analysis, for the results presented in Tables II and III. Without such an analysis, the conclusions must be more tentative than would be case if there had been such an analysis.

You're absolutely correct on this. I didn't include uncertainty in the distances of the individual galaxies in the clusters. There were several reasons. Primarily I figured that since the LEDA database lists the uncertainty in the parameters and therefore should anyone be interested they can check the data uncertainty there. Second, the TFR simply confirmed the already accepted membership of these galaxies within the Virgo cluster. The uncertainty in the data for individual galaxies was not enough to change this.

In the case of the galaxies in Table VII, I did include the uncertainty in the distance modulus because the galaxies were not members of large clusters.

But please proceed with your analysis of the Virgo cluster sample and see if you come up with a different result.

[dgruss23]

Lets talk sensibly, not argue. I wanted to understand your paper, not argue like this. Perhaps this is not the right forum to beg to differ.

Did I question your desire to understand the paper? Of course it's the right forum to "beg to differ". But when you started getting involved in this discussion of the paper back in November with post #2516, you accused me of cherry picking. I wasn't involved at that time and you persisted with that claim into
december. When I got involved in December you persisted in calling it cherry picking despite being shown how the complete sample of 2000 Virgo galaxies reduces to a small number of viable spiral galaxies.

You seem unwilling to ask a question without throwing in an accusation of some sort. Even here implying that my reaction is because you disagree. Quite wrong. My reaction is because you throw out accusations like "cherry picking" and then refuse to acknowledge when your accusation is shown to be wrong.

Then you parachute back after a few weeks away and accuse me of resisting making comparisons.

And now after that behavior you seem to be playing the victim card with your last 2 posts. Whatever. As I said, if you want to ask questions, ask questions. Just don't make accusations that question my integrity.

[Thanatos]

Your integrity is not in question. I did say the data set looks like a cherry pick. I challenged you to refute that assertion. You offered additional selection criteria, but, I am still unable to reproduce your sample set.

What is this 'victim' card thing about? You appear to have confused me for someone who is afraid to ask questions.

[dgruss23]

Your integrity is not in question. I did say the data set looks like a cherry pick. I challenged you to refute that assertion. You offered additional selection criteria, but, I am still unable to reproduce your sample set.

As I explained, the changes in the LEDA database make it impossible to reproduce the exact same sample as the 2005 paper (for which the analysis was actually started in 2003 followed by a 1 1/2 year review process) from the 2006 LEDA database.

However, I have demonstrated in post #2605 using the current LEDA data how the ~ 2000 Virgo galaxies ends up being reduced to such a small number of spirals for Table II. I also provided explanation as to why your list of 92 galaxies is too big.

What is this 'victim' card thing about? You appear to have confused me for someone who is afraid to ask questions.

It has nothing to do with anyone being afraid to ask questions. I'll send you a PM. Why don't we work out our extracurricular bickering in that venue.

[Thanatos]

I found your paper interesting, and worthy of discussion. One of the best I have seen on intrinsic redshift. I disagree with your conclusions, but do not question your integrity.

[VanderL]

I disagree with your conclusions..

Correct me if I'm wrong, but it seems that you (and others) disagree a priori with Russell's conclusions, and will not accept the evidence, no matter what the data analysis shows.
What I'm more interested in is if someone could show the conclusions or analysis of the paper false, and if not (the way it is right now), it means that the conclusions stand, whether you agree or not.

So, imo, it is simply logical to proceed with intrinsic redshift as provisionally acceptable, unless some contrary evidence emerges (directly relevant to Russell's paper) through either the discussion on this board, or papers/arguments from elsewhere. I think it is fair to simply accept intrinsic redshift in spiral galaxies as a placeholder (just like dark matter is a placeholder) and argue from that vantage point.

Cheers.

[TomT]

I think it is fair to simply accept intrinsic redshift in spiral galaxies as a placeholder (just like dark matter is a placeholder) and argue from that vantage point.

I agree with this statement. We have mentioned but not discussed the other galaxy clusters in the Russell paper, which show the same trend as in the Virgo Cluster. So if we start from the vantage point of the data from all these clusters, can we make an estimate of what the numerical value of the intrinsic redshift is? An obvious conclusion is that it varies accordng to galaxy type. Any thoughts on this?
TomT

[Ari Jokimaki]

...can we make an estimate of what the numerical value of the intrinsic redshift is?

I don't think there's enough data for that yet.

An obvious conclusion is that it varies accordng to galaxy type. Any thoughts on this?

Well, if it has to do how tightly the spiral arms are wound, then one could make a case that it's due to compactness, but in that case we might expect Sa galaxies (tightly wound) to have higher redshift than Sc galaxies (loosely wound), if we draw an analogy to gravitational redshifting which is stronger for more compact bodies. But according to Russell's paper it would be Sc galaxies that would have higher redshift. But as it's ScI galaxies which seem to be redshifted most, then perhaps the luminosity class is also a player here. It might be too early days to determine the cause. Russell also discusses possible causes in his paper.

[Thanatos]

I agree I have some 'a priori' objections to the intrinsic redshift conjecture. I think my objections are well founded, but not invulnerable. I also find Russell's papers very interesting and worthy of discussion. I know I rant and rave at him over details, but that is the stuff from which science emerges. I consider him a good scientist. His scholarship and integrity is impeccable. I never for a minute thought he "cherry picked" his data, merely that he left himself open to that accusation. I know I have offended him on that count, and I apologize for that.

[Amber Robot]

If the redshift of galaxies is intrinsic, why does it correlate with distance?

And if the redshift of quasars is intrinsic, how do you explain damped lyman alpha systems?

[TomT]

If the redshift of galaxies is intrinsic, why does it correlate with distance?

Can you give us an example and references to show redshift correspondence with distance. Note that the distance would have to be determined independent of redshift to be a valid indicator of correspondence.
TomT

[Amber Robot]

Can you give us an example and references to show redshift correspondence with distance. Note that the distance would have to be determined independent of redshift to be a valid indicator of correspondence.
TomT

Well, for starters you can look at Hubble's 1929 paper entitled "A Relation between Distance and Radial Velocity among Extra-Galactic Nebulae". That's what started this whole business in the first place.

For others, they are far too numerous to start listing. But I can start listing some examples if you really want me to. The distance/redshift relation is very well established.

Keep in mind that Hubble got the wrong answer for the Hubble constant because they didn't know about the different types of Cepheid variables at the time. The Cepheids were used for the independent distance determination.

[antoniseb]

Can you give us an example and references to show redshift correspondence with distance. Note that the distance would have to be determined independent of redshift to be a valid indicator of correspondence.

What do you consider a valid indicator? Is the trigonometric calculation (parallax, or angular size given a known rate of expansion) the only thing you'll take? How about distance given by the brightness of a type 1a Supernova? How about brightness of the brightest red giants in a galaxy that still is forming new stars? These are all valid methods to me.

[dgruss23]

If the redshift of galaxies is intrinsic, why does it correlate with distance?

This is a thread with a long history, so both your items are things that have been discussed (though not necessarily resolved) before.

The basic model of redshift is not one in which the entire redshift is intrinsic. Instead, there is still a Hubble relation upon which the intrinsic redshift would be superposed.

So the contributions to the observed redshift would be:

Cosmological component - Standard Hubble relation defined by some value of the Hubble constant.

Peculiar velocity component - contribution from orbital motions and bulk flows

Intrinsic component - caused by a mechanism that is unknown at this time.

Normal galaxies would have much smaller intrinsic redshifts than quasars and therefore they produce a Hubble relation. However, the scatter around this Hubble relation may have a significant contribution from the intrinsic component.

And if the redshift of quasars is intrinsic, how do you explain damped lyman alpha systems?

This is another mechanism difficulty. It is obvious that any absorption systems are between the observer and the source that experiences the absorption. However, if the quasars are local these systems would have to be more directly associated with the quasars than the traditionally inferred intervening cosmological distances. For example, Kaminker et al (2000) found that there are maxima and minima in the quasar absorption systems distribution. This type of periodicity - while controversial - has been suggested by Arp et al researchers.

[Amber Robot]

Cosmological component - Standard Hubble relation defined by some value of the Hubble constant.

And is that Hubble relation interpreted as expansion of space? I'm not entirely familiar with these theories, but I thought the point in putting the intrinsic redshift in was to get quasars to fit in with a static universe theory. But I could be wrong.

However, if the quasars are local these systems would have to be more directly associated with the quasars than the traditionally inferred intervening cosmological
distances.

Has Arp et al. addressed the nature of the DLAs in terms of their idea that quasars are local? There's been a lot of studies on DLAs and their accompanying metal absorptions.