favoured topology of Universe candidates

Boud Roukema boud w astro.uni.torun.pl
Śro, 9 Sty 2002, 23:54:48 CET

Dear Ken & Alison,
   I would be pleased for my scientific work to be featured in the
article. :) 

I don't know how much you would like me to write and how much you
would like to rewrite yourselves. But I think it's in the interests of
non-scientists' right to "eavesdrop on science" for me to give you more
detail, including published references, than you are likely to want to
publish. If you wade through the articles and come out feeling dizzy, 
please feel free to ask more questions and I'll try to help clarify 

Even so, I'm not convinced that my initial response explains enough
for the lay reader, but it's already rather long as it stands, so 
have a read through and tell me where you want to go from here.

* General reading:
I would strongly recommend that you read my BASI review


which should be quite accessible to science journalists and give you
my version of how the different approaches relate to one another.

* Affiliation:
By the time your article is published, my affiliation will be
Torun Centre for Astrophysics, University of Nicolas Copernicus, Torun

> 1) Which (if any) of the current candidates for the topology of the universe 
> do you favour, and why?

1a) which candidates do "I favour" ?
1b) why?

1a) which candidates do "I favour" ?

Well, the verb "to favour" is really about my personal pyschological
reactions, which only weakly correlate to scientific reality! 

But I can tell you the three candidates, in chronological order, for
which I think observational predictions should be tested, as well as
the intuitive, subjective probabilities which I've tried to assign to
their being correct - since you asked! If that's what you mean by
"favour", fine.

(1) "three cluster identity": Coma cluster/RX J1347.5-1145/CL09104+4109 
model class: 2-torus
comoving size of Universe: about 1 h^{-1}Gpc (Hubble constant = 100h km/s/Mpc)
subjective probability: 10%

- Serendipitous discovery in discussion of clusters seen in X-rays
http://uk.arXiv.org/abs/astro-ph/9706166 Roukema & Edge    
- optical analysis from archival data
http://uk.arXiv.org/abs/astro-ph/9903038 Roukema & Bajtlik 
- COBE cosmic microwave background (CMB) analysis  (assumes cosmological 
http://uk.arXiv.org/abs/astro-ph/9910272 Roukema   

(2) "COBE counterexample"
model class: 2-torus
comoving size of Universe: about 2 h^{-1} Gigaparsec or greater
subjective probability: 20%

- COBE cosmic microwave background (CMB) analysis  
http://uk.arXiv.org/abs/astro-ph/0007140 Roukema   

(3) "radio galaxy pair": identity of radio-loud active galaxies 3C186/4C+36.21
model class: only one generator hypothesised
comoving size of Universe: about 1 h^{-1} Gigaparsec 
subjective probability: 10%

http://uk.arXiv.org/abs/astro-ph/0111052 Roukema, Marecki, Bajtlik, Spencer

1b) why?

Probably my true answer for (1)-(3) is: because they haven't been falsified
yet. (A technical note: (1) was falsified in astro-ph/9910272 under the
assumption of zero cosmological constant, which was still a reasonable 
assumption at the time the work was done, in 1998, but it is now clearly
a bad assumption, so (1) is again a valid candidate.)

As for the values of my subjective probabilities, well, they're subjective.
I did happen to notice after writing them down that: 

P(subjective)= 10% size of Universe in h^{-1}Gpc  

but this is not a true description of my personal psychological state,
since I would not give a 100% probability to a model 10h^{-1}Gpc in size!

> 2) What research methods are you (and others sharing your view) following, 
> and what progress is being made?

(1) "three cluster identity". 

This candidate makes predictions that it should be possible to point a
telescope at certain angular positions in the sky and detect clusters
at certain distances (redshifts). If a cluster is found at the correct
position and redshift, the hypothesis would be strengthened. If it
were definite that no cluster existed at the predicted position and
redshift, the hypothesis would be falsified.

With my colleagues Jean-Michel Alimi, Jean-Pierre Luminet, Dominique
Proust, (LUTH & GPI, Observatoire de Paris-Meudon), Stanislaw Bajtlik
(Nicolas Copernicus Astronomical Centre, Warsaw) and Etienne
Pointecouteau (Tohoku University, Japan), various requests for
observing time on optical telescopes have been made. We had some
service time allotted in 2001 on the Anglo-Australian Telescope, in
principle, but the schedule did not allow observations to be made. We
will continue this programme of optical telescope tests.

(Note: members of the group differ on the subjective probabilities
they assign to the candidate being the right one!)

(2) "COBE counterexample"

Although this is the candidate I'm probably the most interested in,
it's also the hardest to test, because the larger scale implies that
multiple images ("topological lensed images") of ordinary astrophysical
objects exist at high redshifts, meaning long lookback times into the
past history of the Universe, so that not only is it most difficult to
observe the high redshift images (because they are faint!), it's also 
likely that the objects didn't exist yet at that early epoch.

The MAP and Planck satellites will probably provide the best tests of
this hypothesis, though making a correct analysis with as few as possible
theoretical assumptions is likely to be fiendishly difficult. As far as I
know (but science history is not my field!), Geoffrey Marcy and Paul
Butler missed the detections of exoplanets in their own data in the
early 1990's because they had accepted the theorists' "reasonable"
assumption that massive planets (Jupiters) could not exist in short
period orbits. Such "reasonable" assumptions for detecting cosmic
topology might also lead to a discovery remaining hidden for many

(3) "radio galaxy pair": 

Although the morphologies (shapes) of these two radio galaxy jets are
strikingly similar, the redshift (equivalently, cosmological epoch
when the light was emitted) of only one of them is known.  Radio
galaxy jets expand at very high speeds, from 1% to over 100% (in
apparent speed) of the speed of light.

So, the hypothesis can only be valid if the redshift - and the cosmological
time of emission of the light making the image - of 4C+36.21 is just slightly
higher - cosmological time slightly earlier - than for 3C186. The prediction 
for the redshift of 4C+36.21 is 1.0630 < z < 1.0635. 

A redshift measurement will either strengthen the hypothesis, or falsify it.

My colleague Andrzej Marecki (Torun Centre for Astrophysics, 
University of Nicolas Copernicus, Torun) has time on the William
Herschel Telescope in the Canary Islands for a spectroscopic measurement 
of this redshift, so we hope to have an answer in the coming year!


> From: Kengrimes123 w aol.com
> Date: Mon, 07 Jan 2002 13:18:13 -0500 (EST)
> Subject: Astronomy magazine article.
> To: boud w iucaa.ernet.in

> Dear Dr Roukema

> We are two science journalists currently writing an article for Astronomy 
> magazine (a US title with a circulation of approximately 300,000 readers). 
> The subject of the article is hyperspace and the topology of the universe, 
> and we wonder if you would be interested in being featured in the article.
> The issues we are concerned with are:
> 1) Which (if any) of the current candidates for the topology of the universe 
> do you favour, and why?
> 2) What research methods are you (and others sharing your view) following, 
> and what progress is being made?
> We can be contacted at this return e-mail address (kengrimes123 w aol.com). 
> Should you prefer to speak via telephone, please let us know what would be a 
> suitable time to call you.

Email is fine.

> Many thanks for you attention,
> Ken Grimes & Alison Boyle.

Glad to be of help,

Boud Roukema

boud w astro.uni.torun.pl
Torun Centre for Astrophysics, University of Nicolas Copernicus, Torun
(affiliation by the time the article is published!)


Więcej informacji o liście Cosmo-torun