Farewell Perfect Cosmological Principle?

The standard cosmological principle assumes statistical homogeneity
and isotropy.

A new preprint posted to arxiv.org presents empirical evidence that
this fundamental principle is violated in the messy real world of
nature, as opposed to Platonic idealizations thereof.

http://arxiv.org/abs/1507.00675

Mandelbrot argued that a new and fractal-friendly cosmological
principle might one day be required. Perhaps he was right.

RLO
http://www3.amherst.edu/~rloldershaw

In article ,
"Robert L. Oldershaw" writes:

The standard cosmological principle assumes statistical homogeneity
and isotropy.

Yes and no. A long time ago, people did assume it, and worked out what
things would look like in such a universe. (Some brave souls also
investigated anisotropic models.)

These days, isotropy is an OBSERVATIONAL FACT.

A new preprint posted to arxiv.org presents empirical evidence that
this fundamental principle is violated in the messy real world of
nature, as opposed to Platonic idealizations thereof.

Of course it is violated on small scales. The interesting question in
cosmology is whether at large scales, smaller than the horizon, the
universe is isotropic and homogeneous. This appears to be the case,
despite a handful of people who have been claiming the opposite for
decades.

http://arxiv.org/abs/1507.00675

There, one can read:

The recent discoveries of structures significantly exceeding the
transition scale of 370 Mpc pose a challenge to the cosmological
principle.

This is just wrong. One looks for a scale at which homogeneity and
isotropy dominate, i.e. the scale of the largest structures. This is
the transition scale. Obviously, if one finds something new which is
larger, then the transition scale is larger. As long as it is
significantly smaller than the horizon scale, there is no "challenge".
There is nothing magic about 370 Mpc. It's like discovering a new
species of ape which is bigger than previously known ones: it is new
information, but nothing fundamental. (What WOULD be interesting would
be, say, an ape the size of King Kong, since this would violate
fundamental principles.)

Of course, one can also ask whether "a ring with a diameter of 1720 Mpc,
displayed by 9 gamma ray bursts" is really a structure.

Note that there have been several claims in the literature recently
about the discovery of large objects in the universe, though some of
these "objects" are voids. Note also that almost all of these have been
debunked. Often, bad statistical analyses have been used. There is a
real problem that people call attention to interesting claims, but
hardly anyone cites the papers which debunk them. Just yesterday, I was
reading an article by Jan Conrad in Nature pointing this out.

Mandelbrot argued that a new and fractal-friendly cosmological
principle might one day be required. Perhaps he was right.

Do you have a reference where Mandelbrot talks about cosmology, as
opposed to coastlines or snowflakes?

Le 03/07/2015 19:01, Phillip Helbig (undress to reply) a écrit :
One looks for a scale at which homogeneity and
isotropy dominate, i.e. the scale of the largest structures. This is
the transition scale. Obviously, if one finds something new which is
larger, then the transition scale is larger. As long as it is
significantly smaller than the horizon scale, there is no "challenge".

OK, let's see the facts.

We have a ring of 1.72 GigaParsecs, i.e. 5.6 BILLION light years. This
is more than 40% of the age of the Universe as proposed by current
theory (13.7 Billion years).

Is 40% of the horizon scale *significantly* smaller?

Personally I have some doubts.

With all respect to the professionals here this is an atomic bomb
exploding the LCDM framework. The mass of this behemoth is around 10^18
sun masses!

In principle however you are right Mr Helbig.

"Obviously, if one finds something new which is larger, then the
transition scale is larger."

Sure. Till the next behemoth that will be even larger!

On Friday, July 3, 2015 at 6:14:34 PM UTC-4, jacobnavia wrote:


Sure. Till the next behemoth that will be even larger!

The first predicted "transition to homogeneity" in the modern era
was supposed to occur at about 30 Mpc.

Between then and now it has been repeatedly revised upwards by over
a factor of 10 to about 370 Mpc.

Empirical observations of inhomogeneities on the 1,000 Mpc scale
have been published in the best journals several times in recent
decades.

[[Mod. note -- Perhaps you could clarify which observations you're
talking about? -- jt]]

I would say the writing is on the wall, so to speak.

RLO
Fractal Cosmology

In article , jacobnavia
writes:

We have a ring

9 gamma-ray bursts.

of 1.72 GigaParsecs, i.e. 5.6 BILLION light years. This
is more than 40% of the age of the Universe as proposed by current
theory (13.7 Billion years).

Is 40% of the horizon scale *significantly* smaller?

Yes, because your calculation is wrong. The current distance to the
horizon is more like 40 or 50 billion light years. Due to the expansion
of the universe, you can't just multiply the speed of light by the age
of the universe to calculate the horizon size.

With all respect to the professionals here this is an atomic bomb
exploding the LCDM framework. The mass of this behemoth is around 10^18
sun masses!

9 gamma-ray bursts. A gamma-ray burst is probably a hypernova. So, we
are talking about a few dozen solar masses at most.

Le 04/07/2015 08:35, Robert L. Oldershaw a ecrit :
[[Mod. note -- Perhaps you could clarify which observations you're
talking about? -- jt]]

For a very short review see the cited paper pages 1 and 2. There, they
present data for objects of enormous size discovered in the last years.

Objects that reach the 1240 MPC, i.e. 4.2 BILLION light years...

Read that paper Mr Helbig, it is really mind blowing.

On Saturday, July 4, 2015 at 2:35:39 AM UTC-4, Robert L. Oldershaw wrote:

[[Mod. note -- Perhaps you could clarify which observations you're
talking about? -- jt]]


I have a massive file of reprints on the topic of deviations from "homogeneity" and/or "isotropy". Here are a few recent additions that apply to your question.

Szapudi et al, "Detection of a supervoid...", MNRAS, 2015.

Hutemekers et al, "Alignment of quasar polarizations...", A&A, 2014 0r 2015.

Horvath et al, "Possible structure in the GRB sky distribution..." (structure on the 2,000 to 3,000 Mpc scale), A&A, 2014 or 2015.

Copi et al, "Large-scale alignments from WMAP and Planck", MNRAS, 2013.

....

Going way back, I cannot find the reprint immediately, but I remember that Broadhurst et al published a paper in Nature decades ago on empirical evidence for inhomogeneities on the 1,000 Mpc scale.

Older works are discussed in: http://www3.amherst.edu/~rloldershaw/LOCH.HTM

RLO
Fractal Cosmology

Le 04/07/2015 22:40, Phillip Helbig (undress to reply) a écrit :
In article , jacobnavia
writes:

We have a ring

9 gamma-ray bursts.


That illuminate the ring, whose presence is revealed by the GRBs.


of 1.72 GigaParsecs, i.e. 5.6 BILLION light years. This
is more than 40% of the age of the Universe as proposed by current
theory (13.7 Billion years).

Is 40% of the horizon scale *significantly* smaller?

Yes, because your calculation is wrong. The current distance to the
horizon is more like 40 or 50 billion light years. Due to the expansion
of the universe, you can't just multiply the speed of light by the age
of the universe to calculate the horizon size.


10% then.

It is still 10% of the Universe as proposed by the curreent model!

With all respect to the professionals here this is an atomic bomb
exploding the LCDM framework. The mass of this behemoth is around 10^18
sun masses!

9 gamma-ray bursts. A gamma-ray burst is probably a hypernova. So, we
are talking about a few dozen solar masses at most.


You misunderstood the article. The GRBs ILLUMINATE the behemoth making
it visible!


[[Mod. note -- The article reports observations of the GRBs. The article
does not report any observations of an underlying structure -- that's a
hypothesis requiring further investigation (assuming that the "ring of GRBs"
is in fact confirmed by further analysis & observation).
-- jt]]

It seems to me that there are two key questions in thinking about the
Balazs et al result.

First, is their result truly "surprising"? That is, is their post-hoc
probability estimate 2e-6 valid? Post-hoc probability estimates -- and
their interpretation -- are always a bit dubious, because it's hard to
estimate how many other "equally surprising" patters there might be
which noone has thought to look for.

[If I flip a (fair) coin 10 times in a row, and it
comes up heads every time, that seems rather surprising.
But if you then learn that I'm an insomniac, and do a
dozen or so sets of 10-coin-flippings every night when
I can't sleep, then it's easy to see that I can expect
to get a 10-heads-in-a-row outcome every few months...
so getting one last night wasn't so surprising after all.

In other words, with frequentist statistics, to assess
an experiment we don't just have to know about the
experiment, we also have to know about all the other
things the authors tried that "didn't work"!

This is of course one of the main arguments for taking
a Bayesian perspective. Offhand I'm not sure what a
Bayesian analysis of the Balazs et al input catalog
would look like...]



Second, their result is based on finding 9 "ring" GRBs in a catalog
of 361 GRBs with known redshifts. Since more GRB redshifts are being
measured all the time, if there is truly a physical overdensity of
GRBs in the universe in the region of the sky found by Balazs et al,
then in a few years (when we'll have a much larger GRB-redshift catalog
available) the Balazs et al result should be confirmed in a much more
convincing manner.

And since Balazs et al have pointed to a specific sky region, if we
re-check that specific region with a new GRB-redshift catalog
[I.e., if we re-check using a catalog which does NOT
contain any of the 361 GRBs analysed by Balazs et al]
a few years from now, we won't suffer from the post-hoc-statistics
problem any more.



So, my reaction to Balazs et al is basically "wait and see". If the
result is genuine, within a few years we should have a confirmation
without post-hoc statistics. And if we don't get that confirmation,
that will imply that the result was basically a statistical fluke.

--
-- "Jonathan Thornburg [remove -animal to reply]"
Dept of Astronomy & IUCSS, Indiana University, Bloomington, Indiana, USA
"There was of course no way of knowing whether you were being watched
at any given moment. How often, or on what system, the Thought Police
plugged in on any individual wire was guesswork. It was even conceivable
that they watched everybody all the time." -- George Orwell, "1984"

On Tuesday, July 7, 2015 at 12:40:37 AM UTC-4, Jonathan Thornburg [remove -=
animal to reply] wrote:
=20
So, my reaction to Balazs et al is basically "wait and see". If the
result is genuine, within a few years we should have a confirmation
without post-hoc statistics. And if we don't get that confirmation,
that will imply that the result was basically a statistical fluke.
=20

It is interesting to compare the general responses to empirical
evidence in the case of particle dark matter and the case of
cosmological inhomogeneity/anisotropy.

Forty years of experimental searches have failed to find evidence
for particle dark matter, and yet the general consensus is still
that the dark matter is some kind of subatomic particle.

Over the same period of time there have been published observational
findings that indicate that the inhomogeneity/anisotropy that is
so common on less than cosmological scales continues up to the
largest scales that we can adequately sample. Yet in this case, the
general attitude is to be skeptical of the empirical results and
to assume that the more idealistic models will be vindicated.

Bottom line: How empirical evidence, and the lack thereof, is judged
appears to depend on the answer that is expected on the basis of
prevailing theoretical bias. No surprise there.

RLO
Fractal Cosmology