Monday, September 22, 2008

John Hartnett's Cosmos. 1. Introduction

This is the first entry in what will certainly be a multi-part series of articles on the claims of Australian creationist John Hartnett (see Hartnett's entry in Conservapedia). Since my first piece on Hartnett's work appeared (My Visit to the Creation Forum at McLean Bible Church and a Review of "Hubble, Bubble, Big Bang in Trouble"), I have received a number of inquiries about his work.

In this series, I'll start with a discussion of Hartnett's claims of redshift periodicities. The claim that extragalactic redshifts are distributed in fixed steps, i.e. quantized, is not new. William Tifft of the Steward Observatory at the University of Arizona is probably the originator of the modern redshift quantization movement back in the 1970s. If real, such periodicities or quantization would be very difficult to explain for conventional Big Bang cosmology. This is why you find redshift quantization supported by advocates of other alternative cosmologies such as the Electric Universe (see Electric Cosmos).

I choose to examine the redshift periodicities issue largely because my own graduate work focussed on time-series analysis of very noisy datasets and I accumulated a fair amount of experience working with Fourier series, transforms and power spectra.

Dr. Hartnett has published two papers on the Cornell Preprint server covering the topic of extragalactic redshift periodicities. I'll refer to them as Paper I and Paper II.

  • Paper I: "Galaxy redshift abundance periodicity from Fourier analysis of number counts $N(z)$ using SDSS and 2dF GRS galaxy surveys" by John G. Hartnett, Koichi Hirano (arXiv:0711.4885)

  • Paper II: "Redshift periodicity in quasar number counts from Sloan Digital Sky Survey" by John G. Hartnett (arXiv:0712.3833)


It's interesting that the recently posted third version (v3) of paper I not only has an additional author, but seems to advocate a radically different cosmological model than the second version (v2). In the v2 paper, Hartnett advocated Moshe Carmeli's 5-dimensional cosmological model where the Hubble expansion was made part of the metric. Hartnett published several additional papers based on this model claiming it could explain Dark Matter as well. In paper v3, Hartnett has switched to another model developed by Hirano, Kawabata, and Komiya. This may be because the Hirano et al. cosmology explicitly tries to explain alleged redshift periodicities. This is another reason for me to examine the quantization claims first, as Hartnett appears to be in the process of changing his cosmological model, but quantization is a common component in both of them.

Hartnett is apparently using these works to gain him credibility in the creationist community as a professionally-published cosmologist. In his creationist publications, he has invoked redshift quantization as evidence of Galactocentrism, a feature of his young-universe cosmology. Some graphics similar to those from earlier versions of Hartnett's papers appear in his book Starlight, Time and the New Physics.

Since the advent of the Fast Fourier Transform and the availability of fast personal computers, the ability to compute the power spectral density (PSD) of a dataset has become much easier. Unfortunately, this increased ease of use does not come with an increased understanding of just what the PSD does. In cases with large amounts of low noise data, the PSD can identify well-defined frequencies in the data with relative ease. In other cases, such 'intuitive' understandings of the PSD can easily lead one astray.

Hartnett makes numerous erroneous statements on properties of PSD, suggesting he is relying on his 'intuition' on how the PSD works instead of actually testing the claim. Most researchers, myself included, must demonstrate that our test protocols work for datasets of known content before making such grandiose claims when applying the test to datasets of unknown content. In the abstract for Paper I, Hartnett states that his results "indicate that this is a real effect and not some observational artifact." Yet he has apparently conducted no tests to determine which characteristics of his results are analysis artifacts.

In future entries in this series, I will present an overview of some of Dr. Hartnett's errors in these two papers. In later entries, I'll include some samples of how scientific tests are themselves tested. Some components may be difficult to communicate in this blog since there is limited support for graphics and mathematical notation. Some of these components may take a while to assemble as I may present code snippets (using Python, numpy, scipy, and similar tools) so the reader may explore the analysis themselves. Comments and feedback are certainly welcome.

22 comments:

Rick DeLano said...

If this periodicity is a selection artifact, please address the following claim by Hartnett:
“....there is visible evidence in the raw data for an apparent concentric shell structure centered on the observer.”-- “Galaxy redshift abundance periodicity from Fourier analysis of number counts N(z) using SDSS and 2dFGRS galaxy surveys” J.G. Hartnett
K. Hirano Sep 2008

I can certainly see it in the raw data too:

http://www.sdss.org/includes/sideimages/sdss_pie2.html

W.T."Tom" Bridgman said...

I've actually got a post related to this topic scheduled to appear this weekend.

Two issues:
1) Periodicity/structure: even the cosmological simulations reveal regularity on large scales (which is the topic of the weekend post). However, these are enhancements in the galaxy counts, not strict periodicity.

2) quantization. You see a shell structure? Here's an exercise: Draw the shells you see overlaid on the map. How many galaxies are between those shells? If it's anything greater than zero, it definitely does not indicate a 'quantization'.

Rick DeLano said...

Thank you for your response. It appears that your position is that any observed preferred redshift distributions are insignificant, if they are not absolute.

I do not find this persuasive, and neither does Stephen Hawking:


“...all this evidence that the universe looks the same whichever direction we look in might seem to suggest there is something special about our place in the universe......There is, however, an alternate explanation: the universe might look the same in every direction as seen from any other galaxy, too.....We have no scientific evidence for, or against, this assumption. We believe it only on grounds of modesty: it would be most remarkable if the universe looked the same in every direction around us, but not around other points in the universe.” --Stephen Hawking “A Brief History of Time” 1988 p.42

Now the view of the redshift sky in:

http://www.sdss.org/includes/sideimages/sdss_pie2.htm

provides just that "most remarkable" evidence Hawking lacked in 1988.

Do you honestly think this is irrelevant, simply because you can define "quantization" in a way never employed before in any peer-reviewed use of the term?

I humbly suggest that is not an effective response to Hartnett.

Or indeed to Hawking.

W.T."Tom" Bridgman said...

Tifft interpreted his 'quantization' as sharp, concentric shells and this IS the use in the literature. If you don't mean sharp, concentric shells, then don't use the term 'quantization'. The graphic at SDSS looks nothing like concentric shells.

If you don't believe me, then perhaps you'll believe this guy:
Unknown selection effect simulates redshift periodicity in quasar number counts from Sloan Digital Sky Survey by John Hartnett. He also reports that redshift quantization is not a requirement of his model in an RTB podcast (search for Hartnett).

Rick DeLano said...

Perhaps I can be more precise. Here is my objection: the distribution of redshifts in the SDSS image is not homogeneous.

It is isotropic with respect to the observer, but not homogeneous.

This is contrary to the predictions of the standard model, which predicts both isotropy and homogeneity.

Your citation of Hartnett refers to quasars, not the data I have cited in the SDSS image.

It is worth noting that your own link states that the quasar periodicities clearly *do* exist in the data.

It is also worth noting that Hartnett attributes these periodicities to an "unknown selection effect", certainly not what one would expect were he a shill for a given "creationist" cosmology.

So is Hartnett only right when he suggests he has found evidence against actual periodicities?

Or would he also be right here:

http://arxiv.org/pdf/0711.4885v3

Where he states:

"....galaxies have preferred periodic redshift spacings of ∆z=0.0102, 0.0246, and 0.0448 in the SDSS and strong agreement with the results from the2dFGRS. The redshift spacings are confirmed by the mass density fluctuations, the power spectrum P(z) and N pairs calculations. Application of the Hubble law results in galaxies preferentially located on co-moving concentric shells with periodic spacings."

I am particularly interested in how you handle the mass-density correlations with the redshift periodicities......

W.T."Tom" Bridgman said...

Ah, thank you for reminding me of this.

When the whole redshift quantization thing started with Tifft in the 1970s, it was galaxies, where there were still very small number of samples. As the number of samples grew with larger surveys, the statistical errors declined and the 'quantization' became harder and harder to claim. So supporters had to jump to something with a much smaller number of samples (and larger statistical errors), and quasars were a subset of high-redshift objects that fit the bill.

Of course, this became entangled with H. Arp's discordant redshifts which basically claim that quasar redshifts are not cosmological but due to an intrinsic property. Using this combination means that the 'quantization' has no implications for large-scale cosmological structure.

But suppose you abandon Arp's discordant redshifts? Then you're trying to justify how quasars quantized redshifts should imply control over large-scale cosmological structure if they are a smaller sample than the galaxies. How would quasars respond to this quantization in distance and not galaxies?

The plot you provide at SDSS has reshifts out only to 0.14, when we have samples out to z=4 and further. It also only covers a slice of the sky 2.5 degrees thick in the plane above and below the image. There are plots that cover a broader range of angles and distances
here.

Homogeneity depends on the scale on which you examine.
Testing the Homogeneity of Large-scale Structure with the SDSS Data
The scale of homogeneity of the galaxy distribution in SDSS DR6
Actual tests of the distribution suggest it satisfies homogeneity at large scales.

Hartnett et al. perform a 1-dimensional (radial) power spectrum on a 3-D dataset. The proper analysis is 3-dimensional as noted in my references. In full 3-d, properly accounting for statistical errors, there are no peaks matching Hartnett or similar 'quantization' advocates.
Hartnett never tests his interpretation of the 'peaks'. i.e. Build a simple 3-D cosmological model of galaxies distributed on concentric shells with the spacing claimed. Then process it through his analysis. Do you get similar power spectra? The experiments I have done suggest this will not happen. If I could complete some of my supporting analyses, I hope to write some blog posts on this.

Still, as documented above, Hartnett appears to be backing off on redshift quantization claims since the 2007 paper you reference.

Rick DeLano said...

Thank you for the large scale structure image. It shows precisely the same periodicities shown in the SDSS image I linked earlier.

This would seem to confirm, not refute, Hartnett.

It is important to reiterate: Hartnett does not "back off" his reporting of periodic redshifts. Indeed, he specifically reiterates that they are present in the data.

He simply points to an unknown selection effect as a possible cause *in the case of QSO's*.

Interestingly enough, in the meantime we have a new and very interesting survey of QSO's here:

http://arxiv.org/pdf/1007.2487

Notably:


Thepower spectrum calculated according to Eq. (2) for the whole interval of η (η=0.09 – 1.87) contains the peak at the significance level ex-
ceeding 3σ relatively to the hypothesis of the Poisson η-distribution. In the most statistically representative interval η = 1.148 – 1.870 (z = 1.64 – 4.3) the same periodicity reveals itself at the significance level 4.5σ. Still more prominent peak corresponding to the same period ∆η arises in the power spectrum calculated for the two-
point correlation function ξ (δη) with use of Eqs. (7 –
9)."

He also makes reference to a series of walls and voids, reported first by Broadhurst, and states:

In this context let us note that the correlation function calcu- lated by Mart´ınez et al. (2009) shows a hint of a possi-ble secondary peak at a scale of ∼ 170 h −1 Mpc. Being confirmed such a tendency would be in consent with the hypothesis that at least an intermediate-range or-
der may be present in the spatial distribution of ALSs."

So it appears that other observers are pointing to an extension of these periodicities beyond the 100Mpc level you assert.

So far I see no basis upon which to conclude that Hartnett, Arp, or any of the other researches who have reported non-homogeneous periodic distributions in the redshift sky over the past forty years have any reason to "back off" at all.

Even less so, given the Axis of Evil.

Perhaps it is your own bias against the possibility of a non-Copernican universe that predisposes you one way, just as Hartnett's philosophical preference predisposes him the other?

Imagine that.

It all comes down to philosophical and/or metaphysical preference, just as George Ellis stated decades back........."“People need to be aware that there is a range of models that could explain the observations..... I can construct you a spherically symmetrical universe with Earth at its center, and you cannot disprove it based on observations..... You can only exclude it on philosophical grounds. In my view there is absolutely nothing wrong in that. What I want to bring into the open is the fact that we are using philosophical criteria in choosing our models. A lot of cosmology tries to hide that.”

W.T."Tom" Bridgman said...

Rick,

I released your first post (according to the time tag) suspecting it was the most complete. The error message is a known problem that has yet to be resolved (Blogspot problems posting comments).

In the podcast interview linked above, Hartnett is very non-committal about redshift quantization, which is kind of strange since the podcast date (May 29, 2009) precedes the submission date of the paper (July 24, 2009). But the issue still exists that the power spectrum was computed incorrectly for the type of data, treating it as a 1-D dataset when it is actually a 3-D dataset.

You seem to be confusing structure (patterns in galaxy positions due to collapse of initial inhomogeneities, see Baryon Acoustic Oscillations are NOT 'Redshift Quantization') with periodicities/quantization, which are radically different in the professional literature.

George Ellis and others have made similar statements over the years, but as the amount of data in large-scale surveys increases, it becomes less of a philosophical position and more of a matter of fact.

I would also suggest that the desire to be the 'Center of the Universe', philosophically, has a lot in common with the Sin of Pride.

If you are convinced you see 'periodicities', when you have not done actual measurements on the data, there is nothing that I can say which would convince you otherwise. I've been working for months trying to expand my collection of 'power spectra that can fool you' to prepare a SERIES of blog posts on the topic, as it is so complex. However, if you want to see how just a little noise can afflict a power spectrum, I suggest you check out this excerpt from my Ph.D. dissertation which illustrates how noise impacts time series data. You do know there is a part 2 to this Hartnett article, right?

Rick DeLano said...

Thank you for posting the first "edition", and for the heads up on the error message.

I wish to be as precise as possible in response:

1. In the Large Scale Structure image you linked (not the power series of Hartnett, but the raw data itself), there are very pronounced "clumps" or "shells" of preferred z-values. Do you mean to say that these "clumps" are the result of 500 million light year BAO periodicities?

2. Do you mean to say that these "clumps" are the result of a 2d rendering of a 1d sample of a 3d actuality, and do not actually exist except as an artifact of that dimensional translation?

3. You did not respond to my linked article on QSO periodicities observed at z=1.64- 4.3. The author specifically states that the observed periodicities exist across the entire interval at confidence 4.5. Do you mean to say this author has repeated Hartnett's alleged error, by treating the QSO dataset improperly in his statistical method?

4. Ellis does not agree with you, in terms of your discernment of the difference between scientific fact and philosophical assumption. Neither does Albert Einstein. Obviously, if you are in a position to empirically disprove Ellis' suggested universe, you are also in a position to disprove Relativity. Permit me a prudent skepticism in this regard. I have no doubt, as a philosophical question, that Relativity will be disproven, but I know the difference between a philosophical assumption and an experimental demonstration. The latter, we do not yet possess, as Ellis correctly affirms.

5. The Sin of Pride consists exactly in preferring one's own philosophical preferences, to the Truths by which the Creator has composed the world. Therefore I cordially decline to agree that the Copernican Principle is part of Revelation, and will instead continue to uphold my philosophical preference, while always attempting to keep clearly in mind the difference between scientific proof and philosophical assumption.

6. Thank you for the link to your dissertation. I do intend to review it carefully, but please recall- our exchange is predicated on Hartnett's point that the periodicities exist *in the raw data*.

And they do. See number 1 above.

Thank you for taking the time to continue this most interesting exchange.

W.T."Tom" Bridgman said...

For brevity, I'll answer by the numbers

1 & 2) These are structures are expected from the gravitational collapse. These types of voids and filamentary structures even appear in the formation simulations
N-body simulation, Simulating the joint evolution of quasars, galaxies and their large-scale distribution. There are some structures that might be more suggestive depending on how the full 3-D data is sliced. There's a tool that I've used that lets you 'tour' the SDSS data in 3-D and you see they are more like filaments than regular shells. The tool is demoed in this video.

3) In regards to the Ryabinkovand & Kaminker paper, they use some function, eta(z), with which I am unfamiliar and I don't currently have time to read the paper in detail. However, they do seem to make the same error as Hartnett, treating a 3-D dataset as a 1-D dataset. Overview of the proper techniques are covered in part 2 of the Hartnett article which I linked before. Since a power spectrum of *any* non-zero function has high and low spots, the peaks get interpreted as 'frequencies'. The fundamental theorem of Fourier is that almost *any* function can be constructed as a sum of discrete frequencies and therefore virtually any distribution will contain peaks in the PSD.

4) Ellis' original quote is from a decade that was poor in data. Others have tried to examine other cosmological assumptions but the ones I've heard about had problems when additional data must be used. As for relativity being 'disproven', at worst, it will be shown to be 'incomplete' - just as relativity did not replace Newton, but merely expanded our understanding of the concept of motion and gravity.

5) Historically, astronomy has moved humans further and further from a favored position and my bet is this will continue to be true. I've had very heated e-mail arguments with some fundamentalists over the theological implications. Frankly, Perhaps the universe looks indifferent to humans (a no-win scenario) because it is God's test of our character? See Kobayashi Maru scenario.

6) There is an ambiguity in this description. In Hartnett's 'raw data' there are jumps above the mean level that are reasonably large (Fig 6, Hartnett & Hirano, 2008). But calling them a 'periodicity' is weak and 'quantization' is just wrong since the data doesn't drop to zero galaxies between them.

Right now, you are more interesting than the Electric Universe battle going on in some newer posts. As long as you might bring new ideas to the table that I have not seen before, you have a chance of keeping my interest, depending on my time available, but my response frequency will probably drop soon.

I do have a day job and this is an evening hobby project that I can do as long as other obligations don't intervene. You've also provided me some motivation to work on my demonstrations of erroneous PSD interpretations - and I want time to work on that.

Thanks.

Rick DeLano said...

1 & 2:Thank you for the simulations. The actual lss diagram, by the way, looks quite different from the simulations.

I assume that you do not mean to say that the observable periodicities in galaxy count over redshift in the SDSS lss image is the predicted outcome of an hypothesized app. 148.5 Mpc Baryon Acoustic Oscillation.


But whether you say the periodicities are BAO-related or not, the problem remains: since we observe periodicities in galaxy count/redshift, and since these observations are at variance with (predicted) isotropy and homogeneity, we have evidence of a non-Copernican universe, on at least the scale of the lss image linked.

These periodicities might look like filaments in 3D, they might look like shells or cliumps in 2d, but they represent periodicities in galaxy count over redshift in either case.

They are centered, in the lss image upon the observer.

That is, they are centered upon Earth.

These periodicities would not be observed from a point further removed from Earth (the center of the image), than the preferred periodic z-interval in the distribution.

I find this fascinating.

I invite you to find it fascinating too.


3. Would I be wrong in assuming that your objection would apply to many Fourier transforms performed in the astronomical literature, not necessarily only by Hartnett? And would it be correct that such (flawed) Fourier transforms are used in many other areas on a daily basis including medicine?

I guess my question would be: are you cherry picking here, and objecting to a standard data analysis tool precisely *because* it is yielding results philosophically repugnant to you?

4. Ellis' quote was from 1995. The 90's were poor in data? Hmmm. Since Einstein advanced Relativity in the face of much less data, either poverty is a virtue, or else philosophical preferences determine our cosmological models.

Of course it might be a case of "both/and"......

But Relativity, unlike Newtonian schemes, is based upon an unproven foundational assumption: that "c" is "c" in all reference frames. If physical experiment should prove otherwise, then Relativity falls.

We are only now beginning to be in a position to determine whether the behavior of "c" is as predicted in non-Earth reference frames.

The early evidence is shockingly unsupportive of Relativity (JPL time correction built into GPS software, for example, which renders "c" constant in only one frame. Hint: it ain't the solar system barycenter).

5. SDSS 6 and especially the Axis of Evil move humans directly back into the center of the cosmos, as a matter of *scientific* evidence. Philosophically, of course, we can expect the Copernicans to strive mightily to find a philosophical basis to return Earth to its minor, insignificant status. But the Axis of Evil is a pretty shattering blow. If it stands, then the Copernican Principle falls. Planck should be very interesting in this regard......

6. Preferred periods in galaxy count over redshift accurately describes the anomaly at issue.

Thanks again.......

W.T."Tom" Bridgman said...

1 & 2) I'm not sure what you're describing here, but two issues. In cosmology, homogeneity is defined at a cosmic scale value, a, defined over all space which increases with cosmic time, tau. When we do observations, we see a galaxy at distance, d, we are sampling the cosmos at time tau-d/c. Therefore the density distribution and structure will appear to change with distance due to this sampling effect. The other issue is that if you gave the collection of z values to an experienced signal-processing engineer, they would probably laugh at you for calling those PSD peaks 'periodicities'. Remember, by doing the FFT along the 'redshift axis', a heliocentric bias is automatically installed. Doing a full 3-D power spectrum eliminates that bias. The problems with this get into far more detail than practical for a comment thread and will be part of my post(s) on misuse of the FFT & PSD.

3) There are right ways and wrong ways for applying Fourier transforms to different types of data. I'm sure the medical profession has their standards defined as well and those who do their analysis not to those standards are criticized as well. Hartnett is not the only person who has made these errors. Again, the development and proper application of these techniques are discussed in PART 2.

4) Even 1995 had comparatively sparse galaxy data, though some creationist sites report similar quotes from him as far back as the 1970s.
2dF (2000 SQUARE DEGREE field) with 250,000 galaxies started in 1997, completed around 2000. Twice the size of any previous survey. Mean depth of z=0.1. Covers 10x larger volume than any previous redshift survey. Note survey sky coverage in figure 1.

SDSS didn't start until year 2000
Current sky coverage

Again, your GPS claim has so many problems and is somewhat outside this topic (Hartnett DOES believe relativity is correct), that I can't fill the comment stream with describing the results. I am therefore setting it up as a blog post for the near future.

5) "Axis of Evil": contamination
Low multipoles and other anomalies.

6) Need more evidence than your say-so, especially considering the other impacts it would have.

I'd like to work on these topics in a little more detail which is more appropriate for a blog post where I can include equations and graphics. I'd like to get a few out before November 6, but that will require me to ignore all other comments for a time.

W.T."Tom" Bridgman said...

Note for active comment threads. Priorities of the next few weeks will limit my time on dealing with comments. If there is any comment where I want to respond, it may be significantly delayed. I do have some regular posts queued up to release through the auto-poster.

Rick DeLano said...

1&2: If the peaks are periodic- that is, if the peaks occur at regular intervals (they certainly seem to- see the raw data in your own image)- then of course what would be laughable would be to claim that the periodicities are not periodicities.

I think it might be interesting to survey the redshift literature, and determine what percentage of peer reviewed studies have incorporated the techniques for which you fault Hartnett.

Since you are much more familiar with this, would you care to hazard a guess that the percentage might be....I would be willing to bet the percentage would be "statistically significant".

But the issue remains. A preferred galaxy count over redshift period will be present whether the analysis is 3d or radial.

The question you validly raise, it seems, is how the preferred values would look when mapped- in 2d what might look like be clumps or shells, in 3d perhaps might look more like filaments.

While this is a point well taken, especially since I have a habit of referring to these periodic structures as "shells", perhaps based upon a prejudice instilled by the 2d mapping, I must say nonetheless:

The issue does not resolve the underlying problem for your team, Tom.

Preferred periodic values in galaxy count over redshift.

They are present in the raw data, they are presented again and again and again in the literature.

These periodic structures are explicitly non-Copernican.

They are explicitly geocentric.

Of course there is an "observational bias", since we observe from Earth!

But the observed shells or filaments in the SDSS map would not appear were we to shift our observation point away from the Earth/center by the amount of the observed preferred z-value.

Thus your Copernican dilemma......

3. I'll take your word for it, Tom- and I am not being snarky. It is possible you are right, and have discovered something extremely important. I would never presume to be able to assess the likelihood of this, I only point out that the tool you condemn here is very, very, very widely used in the literature, and it seems that even if you are right, the only result would be to turn shells into filaments, and not to resolve the foundational Copernican dilemma: preferred periods in galaxy count over redshift.

4. There is absolutely nothing in SDSS which would require Ellis to amend a syllable of his 1995 statement. Quite to the contrary, SDSS shows preferred periodic values in galaxy count over redshift, and these periodicities are geocentric. Which enhances the defensibility of the 1995 Ellis quote.

5. It is worth noting that your citation is a 2004 study based on 1 year WMAP data. I suggest instead the 2010 study of the Seven Year data. The news is not good for those who want the Axis of Evil to go away:


"The fact of the alignment is not in doubt, but the significance and implications of the alignment are discussed here....... The probability of two axes randomly aligning in the same
pair of pixels is then 2/196608=0.001%. The probability of getting an alignment within 0.25◦ of a given axis is 0.00095%, which is close to 0.001% above."

Ahem. That is, 99.999 percent.

I believe the term is "statistically significant".

If the Axis of Evil stands, then the Copernican principle falls.

And as of right now it looks mighty good that the Axis of Evil will stand.

SEVEN YEAR WILKINSON MICROWAVE ANISOTROPY PROBE (WMAP1) OBSERVATIONS:
ARE THERE COSMIC MICROWAVE BACKGROUND ANOMALIES?
http://arxiv4.library.cornell.edu/pdf/1001.4758v1

PS: Tom, I only now noticed your post on the Geocentrism Conference. I have sent you a note there.

W.T."Tom" Bridgman said...

It doesn't matter how many people are using the wrong methodology. That does not make it correct.

It would be nice if there were a comprehensive tutorial with detailed demonstrations of why and how such sampling produces these artifacts, but I have yet to find one. Most of the researchers have to discover it for themselves. For most of those reporting periodicities with the radial technique, today you find many of them do most of their work in areas other than cosmology - so they don't take the time to learn the nuances of the technique. They have a MatLab or IDL toolbox that generates an FFT and they just blindly use it without experimenting with what their process does to signals of KNOWN content.

Clearly you did not understand the time series writeup from my Ph.D. Figure 1 demonstrated how much the FFT could distort the original input signal and figure 2 is an example where the input signal with a KNOWN POWER SPECTRUM is convolved with some random noise. The simple single-pass FFT, which is often used when analyzing on the single z-axis, generates peaks all along the band - but they are all manifestations of noise propagated through the FFT. The bottom panel of Figure 2 shows how segmenting the original time series and summing them reveals the PSD of the input signal. Of course, everyone doing this in the single z-axis probably don't like to do that because - surprise - all their nice peaks disappear!

Rick DeLano said...

Tom: Have you ever demonstrated in any way at all, that Hartnett (or the other, scores and scores of peer reviewed scientific studies which employ Fourier analysis) have failed to test and secure their analyses against the flaws you point out?

You certainly have not done so in the case of Hartnett.

You in fact have not done so in the case of any of the hundreds of peer reviewed papers which have incorporated Fourier analysis in examining the question of periodicities in preferred galaxy count over redshift.

Now that the Conference is over (and what a tremendous success it was!) I shall attempt to find the time to examine your criticism, as it involves a shocking allegation of incompetence, over decades and longer, on the part of every scientist employing Fourier analysis on galaxy surveys...................

I'm not sure which is more shocking- to learn that the predicted isotropic and homogeneous Universe of the mainstream Relativity theory does not exist, as can be ascertained by anyone who views the SDSS images linked above; or that essentially every scientist working in this field who has employed a Fourier analysis is a blithering incompetent....

Are you certain this is the case you really want to make, Tom?

W.T."Tom" Bridgman said...

Quote:Tom: Have you ever demonstrated in any way at all, that Hartnett (or the other, scores and scores of peer reviewed scientific studies which employ Fourier analysis) have failed to test and secure their analyses against the flaws you point out?

You certainly have not done so in the case of Hartnett.

You in fact have not done so in the case of any of the hundreds of peer reviewed papers which have incorporated Fourier analysis in examining the question of periodicities in preferred galaxy count over redshift.

*I* do not have to. Numerous others, in previous years have demonstrated the correct way to do these power spectra. Considering there are a handful of well-documented correct ways and potentially an infinite number of wrong ways, why should researchers spend their valuable time correcting the errors of those who don't take the time to understand the behavior of the tools they are using, especially when the details are so readily available?

Nonetheless, the tutorial will begin soon.

Quote:Now that the Conference is over (and what a tremendous success it was!) I shall attempt to find the time to examine your criticism, as it involves a shocking allegation of incompetence, over decades and longer, on the part of every scientist employing Fourier analysis on galaxy surveys...................

I'm not sure which is more shocking- to learn that the predicted isotropic and homogeneous Universe of the mainstream Relativity theory does not exist, as can be ascertained by anyone who views the SDSS images linked above; or that essentially every scientist working in this field who has employed a Fourier analysis is a blithering incompetent....

Are you certain this is the case you really want to make, Tom?

Considering that I have found no evidence that you have exhibited any competence in this field, while claiming absolute knowledge, you are in a very poor position to make such a statement. I rank those who make a real effort, but make some mistakes, in far higher regard than those who make no effort in a subject yet claim expertise.

Rick DeLano said...

Hmm. This last post is certainly a change in tone, Tom. I note that you have, for the first time, departed in a very noticeable way from accurately characterizing my words.

Why?

You say:

Considering that I have found no evidence that you have exhibited any competence in this field,

>>" In fact I have openly admitted this, in the post of Oct 18 above:

"I think it might be interesting to survey the redshift literature, and determine what percentage of peer reviewed studies have incorporated the techniques for which you fault Hartnett.

***Since you are much more familiar with this***, would you care to hazard a guess that the percentage might be....I would be willing to bet the percentage would be "statistically significant".

But the issue remains. A preferred galaxy count over redshift period will be present whether the analysis is 3d or radial."
*********

while claiming absolute knowledge,

>>I have never made any such claim. It is not conducive to good debate to mischaracterize one's interlocutors' words.

Up until now I had thought this drearily typical tactic would have been beneath you.
********************

you are in a very poor position to make such a statement. I rank those who make a real effort, but make some mistakes, in far higher regard than those who make no effort in a subject yet claim expertise.

>> Tom, you still have yet to answer the question. How are all these power series papers showing preferred periodic redshift values getting published in peer reviewed journals, if it is all just a matter of easily demonstrated faulty procedures?

Why hasn't the peer review process caught this?

Why haven't you published a peer reviewed analysis of specific instances of these flawed techniques?

Or at least published a specific answer to my questions:

1. What evidence do you have that Hartnett, or any of the many other authors reporting periodic preferred values in redshift/galaxy count, have in fact made the errors you set forth in your primer?

So far we have seen none.

2. Even if you were to show Hartnett and others guilty of the basic incompetence you allege (but have yet to demonstrate), this does not address the statement of Hirano and Hartnett, that the raw data (*not* the power series) shows visible evidence of a concentric shell series centered upon the observer (Earth). How could a flawed power series be responsible for periodicities in the raw data?

You say:

why should researchers spend their valuable time correcting the errors of those who don't take the time to understand the behavior of the tools they are using, especially when the details are so readily available?

>> But Tom, I thought that was exactly the *point* of peer review and challenges to published studies?


If you have reached the end of your civility and sense of fair play, and would prefer to have recourse to the argumentum ad hominem, then be my guest, Tom.

I will be delighted to address these questions to others if you are unable or unwilling to answer them.

Anonymous said...

Dear Mr. Bridgman,

it's strange to see that Rick DeLando has the last word about such a fringe theory.
Could you please give an answer to his post, especially the following three statements:

"A preferred galaxy count over redshift period will be present whether the analysis is 3d or radial."

"1. What evidence do you have that Hartnett, or any of the many other authors reporting periodic preferred values in redshift/galaxy count, have in fact made the errors you set forth in your primer?"

"2. Even if you were to show Hartnett and others guilty of the basic incompetence you allege (but have yet to demonstrate), this does not address the statement of Hirano and Hartnett, that the raw data (*not* the power series) shows visible evidence of a concentric shell series centered upon the observer (Earth). How could a flawed power series be responsible for periodicities in the raw data?"

Kind regards, and keep up the good work!

W.T."Tom" Bridgman said...

It is fairly straightforward to show that a 3-D power spectrum cannot be the same as a radial power spectrum if the angular position information is ignored. Doing a PSD of the radially-distributed galaxy counts IS NOT mathematically equivalent to the 3-D power spectrum.

PSD(f(x,y,z)) = PSD(f(r,theta,phi)) != PSD(f(r))

since there are slightly different numbers of galaxies in different directions (they are not 100% spherically symmetric around the Earth). I'll also try to assemble an example case illustrating these for a future post.

If you read carefully, you'll see that I don't make that accusation against Hartnett. Hartnett and others have made an analysis error (I make those myself). The issue becomes if he and others learn from the mistakes and how to correct them.

However, I see it has been two years since I last worked on this problem. I checked my codes for this project and their still working and I have some ideas on how to resolve the problems that hindered progress (see Quantized Redshifts XI. My Designer Universe Meets Some Data and What's Next...).

Mark said...

Tom:

The Planck data has been released, and the axis of evil is still there. Planck used absolute thermography, and a different scanning pattern in an attempt to remove any potential artifacts, yet the "artifacts remain.

It looks like the Copernican Principle has taken another hit.

W.T."Tom" Bridgman said...

To Mark,

I've downloaded and perused the Planck papers but not yet had a chance to read them in detail.

Remember that the CMB signal is still determined by taking the total received signal and subtracting known foreground signals. What remains is interpreted as the CMB.

But as detector sensitivity is improved, it becomes a real issue as to whether signals from fainter foreground sources must be subtracted. Even the Planck papers mention contributions from the *local* Solar System that may need to be included. These local contribution could explain the anomaly without problems for the Copernican Principle.

Tom

So...What Happened?

Wow.  It's been over eight years since I last posted here... When I stepped back in August 2015,...