The Supervoid and the Cold Spot

While I was away at the SEPnet meeting yesterday a story broke in the press broke about the discovery of a large underdensity in the distribution of galaxies. The discovery is described in a paper by Szapudi et al. in the journal Monthly Notices of the Royal Astronomical Society. The claim is that this structure in the galaxy distribution can account for the apresence of a mysterious cold spot in the cosmic microwave background, shown here (circled) in the map generated by Planck:


I’ve posted about this feature myself here in the category Cosmic Anomalies.

The abstract of the latest paper is here:

We use the WISE-2MASS infrared galaxy catalogue matched with Pan-STARRS1 (PS1) galaxies to search for a supervoid in the direction of the cosmic microwave background (CMB) cold spot (CS). Our imaging catalogue has median redshift z ≃ 0.14, and we obtain photometric redshifts from PS1 optical colours to create a tomographic map of the galaxy distribution. The radial profile centred on the CS shows a large low-density region, extending over tens of degrees. Motivated by previous CMB results, we test for underdensities within two angular radii, 5°, and 15°. The counts in photometric redshift bins show significantly low densities at high detection significance, ≳5σ and ≳6σ, respectively, for the two fiducial radii. The line-of-sight position of the deepest region of the void is z ≃ 0.15–0.25. Our data, combined with an earlier measurement by Granett, Szapudi & Neyrinck, are consistent with a large Rvoid = (220 ± 50) h−1 Mpc supervoid with δm ≃ −0.14 ± 0.04 centred at z = 0.22 ± 0.03. Such a supervoid, constituting at least a ≃3.3σ fluctuation in a Gaussian distribution of the Λ cold dark matter model, is a plausible cause for the CS.

The result is not entirely new: it has been discussed at various conferences over the past year or so (e.g this one) but this is the first refereed paper showing details of the discovery.

This gives me the excuse to post this wonderful cartoon, the context of which is described here. Was that really in 1992? That was twenty years ago!

Anyway, I just wanted to make a few points about this because some of the press coverage has been rather misleading. I’ve therefore filed this one in the category Astrophype.

First, the “supervoid” structure that has been discovered is not a “void”, which would be a region completely empty of galaxies. As the paper makes clear it is less dramatic than that: it’s basically an underdensity of around 14% in the density of galaxies. It is (perhaps) the largest underdensity yet found on such a large scale – though that depends very much on how you define a void – but it is not in itself inconsistent with the standard cosmological framework. Such large underdensities are expected to be rare, but rare things do occur if you survey a large enough volume of the universe. Large overdensities also arise as statistical fluctuations in large volumes.

Second, and probably most importantly, although this “supervoid” is in the direction of the CMB Cold Spot it cannot on its own explain the Cold Spot; the claim in the abstract that it provides a plausible explanation of the cold spot is simply incorrect. A void can affect the measured temperature of the CMB through the Integrated Sachs-Wolfe effect: photons travelling through such a structure are redshifted as they travel through the underdense region, so the CMB looks cooler in the direction of the void. However, even optimistic calculations of the magnitude of the effect suggest that this particular “void” can only account for about 10% of the signal associated with the Cold Spot. This is a reasonably significant contribution but it does not account for the signal on its own.

This is not to say however that it is irrelevant. It could well be that the supervoid actually sits in front of a region of the CMB sky that was already cold, as a result of a primordial fluctuation rather than a line-of-sight effect. Such an effect could well arise by chance, at least with some probability. If the original perturbation were a “3σ” temperature fluctuation then the additional effect of the supervoid would turn it into a 3.3σ effect. Since this pushes the event further out into the tail of the probability distribution it makes a reasonably uncommon feature look  less probable. Because the tail of a Gaussian distribution drops off very quickly this has quite a large effect on the probability. For example, a fluctuation of 3.3σ or greater has a probability of 0.00048 whereas one of 3.0σ has a probability of 0.00135, about a factor of 2.8 larger. That’s an effect, but not a large one.

In summary, I think the discovery of this large underdensity is indeed interesting but it is not a plausible explanation for the CMB Cold Spot. Not, that is, unless there’s some new physical process involved in the propagation of light that we don’t yet understand.

Now that would be interesting…


9 Responses to “The Supervoid and the Cold Spot”

  1. Dear Peter,
    Thanks for calling out the incorrect claim in the abstract re the ISW effect. I’m amazed at the standard of refereeing that can allow such a paper to be published, given that it has in effect been “pre-butted”: a paper we published last November already showed that this claim was nonsensical.

    I’d also like to point out that this so-called “supervoid” is in fact not even “the largest underdensity found on such scales”. I’m told Roberto Trotta (who is quoted in the Guardian piece) was making the same claim on Twitter but it simply isn’t true. Voids up to 1.3 times larger and with a central density approximately 4 times smaller have already been seen in other galaxy data (here, start of Section 6.3 for the sizes), and in simulation (here). Both from simulation and from theory one would expect ~20 such voids in our neighbourhood, so it’s not very unusual. I’m even told that Szapudi et al themselves found an even bigger void in the WISE-2MASS data, but it wasn’t anywhere near the Cold Spot so they left it out of the paper.

    Oh and an aside – it’s true that one definition of a “void” is a region completely empty of galaxies, but it isn’t the only definition. It probably isn’t even the most useful definition.

    • Phillip Helbig Says:

      As it happens, I’ve been involved in several conversations (publicly and privately) recently about the quality (or lack thereof) of refereeing and how to improve it. One suggestion I’ve heard a few times deserves, I think, serious consideration: After a paper is accepted, the name(s) of the referee(s) are published along with the paper. This wouldn’t violate any legitimate reason to keep refereeing anonymous (if the referee so desires) before acceptance (which, of course, might never happen), but should force people to be a bit more careful and, if there is a goof, accept part of the blame.

      There used to me something similar in MNRAS: After the author list, it would say “Communicated by XXX”.

      • Phillip Helbig Says:

        Just to avoid any confusion, especially since (s)he might be reading this, this does most definitely not apply to my one currently submitted paper, where I awaiting a report on the revised version. In this case, the referee’s response was prompt and the comments were helpful. 🙂

    • Phillip Helbig Says:

      Readers might want to check out some more discussion on this.

  2. Your calculation of probabilities leaves out an important effect though. As you say, maybe a supervoid sitting in front of a “3 sigma” perturbation on the last scattering surface could make it look like a “3.3 sigma” anomaly via the ISW effect (I’d actually dispute this quite strongly, especially for this void).

    But what is the total probability of the (a) existence of the original “3 sigma” anomaly, (b) the existence of the supervoid, and (c) the alignment of said supervoid and the existing anomaly on the sky? In fact (we showed in the paper above) it is always less than or equal to the probability of a “3.3 sigma” anomaly in the first place.

    In other words, no supervoid can explain the Cold Spot.

  3. Han Vinke Says:

    These Rees-Sciama fluctuations are normally too small to be solely responsible for the Cold Spot.

    They should be carefull jumping to conclusions in my opinion.

    Another good reason for instance could be inverse Compton scattering. There is the possibility that fast relativistic electrons can’t escape from the omnipresent background radiation of the CMB.
    A part of the low energy photons would be upscaled to much higher energies, turning into x-ray or even gamma-ray photons.
    This way they would be invisible for the limited bandwidth view of PanSTARRS-1 and Wise.
    It would result in a synchro-Compton catastrophe.

    Within the same Cold Spot region there is a known soft X-ray “Hot Spot” a.k.a. the Orion-Eridanus Superbubble. It is full of ionised hydrogen (and free electrons) creating Hα emitting filaments.

  4. Phillip Helbig Says:

    “Was that really in 1992? That was twenty years ago!”


    The Moving Finger writes; and, having writ,
    Moves on: nor all your Piety nor Wit
    Shall lure it back to cancel half a Line,
    Nor all your Tears wash out a Word of it

  5. […] the time in sky patterns simulated using the model assumptions. One possible explanation of this ( which I’ve discussed before) is that this feature is generated not by density fluctuations in the primordial plasma (which are […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: