Astronomy Without A Telescope – The Edge Of Significance


Some recent work on Type 1a supernovae velocities suggests that the universe may not be as isotropic as our current standard model (LambdaCDM) requires it to be.

The standard model requires the universe to be isotropic and homogeneous – meaning it can be assumed to have the same underlying structure and principles operating throughout and it looks measurably the same in every direction. Any significant variation from this assumption means the standard model can’t adequately describe the current universe or its evolution. So any challenge to the assumption of isotropy and homogeneity, also known as the cosmological principle, is big news.

Of course since you are hearing about such a paradigm-shifting finding within this humble column, rather than as a lead article in Nature, you can safely assume that the science is not quite bedded down yet. The Union2 data set of 557 Type 1a supernovae, released in 2010, is allegedly the source of this latest challenge to the cosmological principle – even though the data set was released with the unequivocal statement that the flat concordance LambdaCDM model remains an excellent fit to the Union2 data.

Anyhow, in 2010 Antoniou and Perivolaropoulos ran a hemisphere comparison – essentially comparing supernova velocities in the northern hemisphere of the sky with the southern hemisphere. These hemispheres were defined using galactic coordinates, where the orbital plane of the Milky Way is set as the equator and the Sun, which is more or less on the galactic orbital plane, is the zero point.

The galactic coordinate system. Credit:

Antoniou and Perivolaropoulos’ analysis determined a preferred axis of anisotropy – with more supernovae showing higher than average velocities towards a point in the northern hemisphere (within the same ranges of redshift). This suggests that a part of the northern sky represents a part of the universe that is expanding outwards with a greater acceleration than elsewhere. If correct, this means the universe is neither isotropic nor homogeneous.

However, they note that their statistical analysis does not necessarily correspond with statistically significant anisotropy and then seek to strengthen their finding by appealing to other anomalies in cosmic microwave background data which also show anisotropic tendencies. So this seems to be a case of looking at number of unrelated findings with common trends – that in isolation are not statistically significant – and then arguing that if you put all these together they somehow achieve a consolidated significance that they did not possess in isolation.

More recently, Cai and Tuo ran much the same hemispherical analysis and, not surprisingly, got much the same result. They then tested whether these data favoured one dark energy model over another – which they didn’t. Nonetheless, on the strength of this, Cai and Tuo gained a write up in the Physics Arxiv blog under the heading More Evidence for a Preferred Direction in Spacetime – which seems a bit of a stretch since it’s really just the same evidence that has been separately analysed for another purpose.

It’s reasonable to doubt that anything has been definitively resolved at this point. The weight of current evidence still favours an isotropic and homogeneous universe. While there’s no harm in mucking about at the edge of statistical significance with whatever limited data are available – such fringe findings may be quickly washed away when new data comes in – e.g. more Type 1a supernovae velocity measures from a new sky survey – or a higher resolution view of the cosmic microwave background from the Planck spacecraft. Stay tuned.

Further reading:
– Antoniou and Perivolaropoulos. Searching for a Cosmological Preferred Axis: Union2 Data Analysis and Comparison with Other Probes.
– Cai and Tuo. Direction Dependence of the Deceleration Parameter.

16 Replies to “Astronomy Without A Telescope – The Edge Of Significance”

  1. Yo Steve, I have to question your usage of the term “homogenous”. According to the Oxford Dictionaries – Homogeneous; Usage:

    The usual spelling is homogeneous, and the spelling homogenous is traditionally regarded as an error. Homogenous is a different word, a specialized biological term meaning ‘having a common descent’, which has been largely replaced by homologous. From the evidence of the Oxford English Corpus, the spelling homogeneous has become significantly less common since 2000, and around a third of citations for the word now use the form homogenous. This can now be regarded as an established variant.

    However, I think that Universe Today should maintain standards and not sink to the level of lesser journals.

    1. Besides, Homogeneous has, at least phonetically, the word genius in it. And we’d like to think they’re scattered evenly about the heavens as well! (: Like here in the forum.

    2. Spelling is only an issue where it impedes reading–which these typos do not. Shakespeare we should remember never on the six occasions of writing his name spelt it “Shakespeare” but “Shaksper” or “Shakspere”. Correct spelling is by itself not that important.

      I would much prefer the occasional spelling “error” and a steady stream of quickly written and informative pieces than fewer but word perfect ones.

      1. I agree with your point, more great articles with a few typos are okay, unfortunately this is the way most children are educated these days. (Not you, Steve) Phonetics works in some instances, but scientific venues are not the right ones. I think the problems stem from teachers themselves who cannot spell.
        Otherwise, another fine and informative article at UT standards.

  2. I almost wish this had not been posted here. I suppose I may have to read their paper in some detail. I recognize some computations in it, so these people appear to have a handle on the subject. The lack of reporting on this 9 months old paper, at least as far as I have seen, suggests this has not generated much concern.

    Physically it makes little sense that the universe would have anisotropy on the largest scales observable. First off the CMB does not display this sort of feature, which means that if these observations are correct they reflect a more local feature. If such an anisotropy of the nature exist on the largest scale observable it would mean there is some net generation of momentum in some particular direction. If something like that were to exist it would mean the universe simply makes no sense at all.

    The correct spelling is homogeneous. The word homogenous is a shortened version that came from the dairy industry.


    1. I don’t mind if you don’t read the paper 😉

      BTW – some ivan3man-inspired research informs me that any milk-related usage should also use the homogeneous spelling.

      The spelling confusion may arise from the verb ‘to homogenize’ – which doesn’t employ the extra e (so homogenized milk is homogeneous).

      1. The paper looks like an honest analysis. This could tell us something about the scale of anisotropies. Of course this is a possible intellectual obligation on top of other things. I don’t thing there is some cosmic preferred direction in a spatial direction. There is of course a temporal preferred direction, with interesting questions on the meaning of energy.

        In the end it is a question of God vs the devil. To take the side of God is to say there are great patterns or symmetries to the universe. The devil tries to tell us there is no such thing. I tend to prefer to God view on this, but the devil has some cards up his sleeve — and we can’t prove him wrong in the end.


      2. I don’t question the paper’s honesty – but think that the weak statistical significance does not enable strong conclusions. More data will either support or refute this claim. That’s science – but I thought the Arxiv blog was over-stating the situation.

        Not sure how helpful the deity analogy is. An asymmetric universe is more likely the result of a mysterious agency that is pulling levers to make it work, while a symmetric universe has no obvious evidence of a controller (from either side of the fence).

      3. It is the natural assumption that such an asymmetry would be due to some underlying foundation. The question of course is whether we could access data on such a substratum. We are entering a stage in physics where our observations of both the quantum small domain of particles and the deepest reaches of the cosmos are becoming more indirect. It poses the prospect that before long our understanding of physical or cosmic foundations may become lost. It is my hope that we can come to some effective theory, a grand unification or quantum cosmology, which works well enough up to what ever limits we have to our abilities to observe. This is even given the likelihood there are deeper foundations which we may never be able to access.


    2. “The correct spelling is homogeneous. The word homogenous is a shortened version that came from the dairy industry.” Which version then is applicable when speaking specifically about OUR GALAXY ONLY?? …wait for it…. Ya…

  3. Sorry for this layman question but can’t this difference be a result of relative Milky Way movement?

  4. Sorry for this layman question but can’t this difference be a result of relative Milky Way movement?

    1. Hi Marcin – interesting point. Off the top of my head, I think it’s unlikely the (very big) galaxy would shift sufficiently within the time period that the measurements were taken to affect the measurements (notwithstanding I am not convinced that a significant difference has been demonstrated anyway).

      If galaxy movement was significant there should be a Doppler shift effect. I don’t know that the researchers in any way sought to control for this – so perhaps they also assumed the time period was long enough for such an effect to be significant. But given the tiny alterations they appeal to for representing significance – perhaps they should have.

  5. Well, extraordinary claims needs extraordinary evidence. The low redshifts used (< 1) would take us out something like 2 Gpc. So that is a 1/10th of the observable distance (~ 14 Gpc radius).

  6. As usual LC is correct on all counts. Isometry would have been the correct word instead of anisotropy. Heehheh nice one LC on the diary industry term.

    That’s the question to ask concerning the CMB question and current observations but at least the author was attempting to distill a complex paper for the masses… Even mediocre science articles is better than none.

    Thanks to LC and Universe Today – Keep the articles [and funny blog comments] coming – I love the stuff 🙂



Comments are closed.