Neutrinos on Speed

The internet, twitterdom, blogosphere, and even the mainstream media are all alive today with wild speculations about a curious claim that neutrinos might travel faster than light.

If you’re interested in finding the source of this story, look at the arXiv paper here. I haven’t got time to go through the paper in detail, but I think it must be an instrumental artefact or some other sort of systematic error.

One major reason for doubting the veracity of the claim that neutrinos travel faster than light is provided by astronomical observations. Neutrinos produced by the explosion of Supernova SN1987a were detected when it went boom in 1987, approximately three hours before the visible light from SN 1987A reached the Earth.

The few hours delay between neutrinos and photons is explained by the fact that neutrino emission occurs when the core of the progenitor star collapses, whereas visible light is released only when a shock wave reaches the surface of the imploding object. Three different experiments detected (anti)neutrinos: Kamiokande II found 11 , IMB 8 and Baksan 5, in a burst lasting less than 13 seconds.

If the time delay reported by the OPERA detector over the distance between CERN and Gran Sasso were extrapolated to the distance between Earth and SN1987a then the neutrinos should have arrived not a few hours early, but a few years, and there would not have been coincident arrivals at the different detectors on Earth.

Do neutrinos go faster than light?
Some physicists think that they might.
In the cold light of day,
I am sorry to say,
The story is probably shite

UPDATE: Now that I’ve read the paper let me point out that the OPERA result is essentially

δv/c = (2.48 ± 0.28(stat) ± 0.30(syst)) × 10-5,

whereas the constraints from Supernova 1987a work out to be   δv/c < 2 × 10-9 for  neutrino energies of 10 MeV. See the comments below for discussion.

I’ll also mention at this point that the analysis done in the paper is entirely based on frequentist statistics. Somebody needs to do it properly.


29 Responses to “Neutrinos on Speed”

  1. Oh well. At least it took our minds off the UARS satellite and collapse of the western economy for a couple of hours.

  2. The neutrinos in the OPERA experiment are passing along a path over 700km in length — meaning that halfway along their path, they’re 0.7% closer to the center of the Earth than the point of emission or reception. I need more details about how they aimed the beam, but I can tell you that accounting for this difference incorrectly would introduce an error of around 10^-3 in timing and I can’t see anywhere in the OPERA paper where they mention anything like this.

  3. I suspect some researchers are going to experience embarrassment over this issue, and I don’t expect it will include anyone backing the conventional interpretation of relativity.

    Perhaps somebody has made an error in a calculation of the distance between source and detector, or there has been a slip in measuring time differences?

    • Rhodri Evans Says:

      The distane error is my own hunch. They claim the distance is measured to an accuracy of 20cm. But how? I’ve heard via Twitter it was using GPS. I can’t believe GPS could give this level of accuracy over 738km.

  4. Hmm. So this is an example of “community peer review” that you propose is it? An admission you haven’t read the paper in detail but a declaration that its “probably shite” anyway. Not a very nice way to treat fellow scientists (who themselves declare the SN evidence against their result) in my view.

    • This is a personal blog, not community peer review.

      Also the limerick was (a) about the “story” (i.e. the press reaction) rather than the paper and (b) intended to be humorous.

      PS. I wouldn’t have used the word “shite” if I could have found a rhyme for “bollocks”…

      • Hi Peter,

        Of course you are correct this is a personal blog, and you are free to say whatever you like.

        I just worry however that in the community peer review model that you are keen on, the confusion between “opinion” and “peer review” will get muddled in some peoples minds (I am not accusing you here!) even on the “official” website that hosts the reviews.

        My concern is that the kind of “coffee time/pub” discussion of a topic (which is perhaps a good model for what this blog represents) becomes the only scrutiny a piece of work gets. A result that conflicts with someones own current view of the subject will get shot down not with an engagement with the evidence they have presented (which is peer review in my view) but with general statements about how it can’t possibly be true and they must have got it wrong somehow.

        Anyway, I fear I have dragged us somewhat off topic here…


    • I think Peter was referring to comments on “normal” papers. In this case, it seems that the folks are saying “we don’t believe it ourselves, but have a look and help us find what we must have overlooked”. So, the authors themselves think it is shite.

  5. Todd Laurence Says:

    The only scientific verification of psychic
    relativity, i.e., the ability of ‘mind’ to
    transcend space and time, per acausality
    theory-Professor W. Pauli, Nobel laureate,
    neutrino particle:

    • Anton Garrett Says:

      Actually there’s decisive evidence of acausality, since the order of measurement of the spin of the two particles in Bell expts is not Lorentz-invariant. But there is no evidence at all that the human mind can influence the outcome.

      • I agree with what you say about acausality in quantum physics, but don’t the experiments to which you refer require information velocities >>c?

      • Anton Garrett Says:

        The results of Bell-test experiments force nonlocality, and also force superluminality if you try to maintain causality. But if you renounce causality then you don’t need superluminality; and it seems to me that you are pretty much forced to acausality anyway from the fact that the order of the measurements on the 2 particles is not Lorentz-invariant.

  6. My take on this is actually that the authors have been commendably open about the whole thing. I fully agree that this is almost certainly some systematic uncertainty that they haven’t been able to identify (and the paper was less thorough than I’d have liked when it came to the error analysis).

    However it is less clear whether the SN1987A results (which they quote in the introduction of the paper) are relevant. That was for electron anti-neutrinos and the ones here are muon/tau (not entirely clear to me) neutrinos, and the energy difference is significant – all of which doesn’t matter within the current model but if something broke down it wouldn’t add a lot to the incredulity to introduce some energy/particle dependent velocity effect.

    Anyway, almost certainly much ado about nothing and xkcd has captured it perfectly 🙂

    • Sorry for repeating your point below. I should have read more carefully before commenting.

    • Yes, it’s true but remember that neutrino flavour states are not energy eigenstates; we know there’s oscillation between e and mu flavours. There may be a way of reconciling neutrino oscillations with an energy-dependent effect but I’m not aware of it. It would presumably require one or of the neutrino energy states to be superluminal.

      In any case I was just saying that the SN neutrinos make me very suspicious of this result, not that they disprove it.

    • Yes, I completely agree with that statement. But I’m a pedant, and couldn’t resist being pedantic.

  7. believe that for a moment. The experiments also imply you can send a message back in time, but I don’t

  8. Not that I believe the result, but … the SN1987a argument is not really watertight, is it? As mentioned in several places, the SN1987a neutrinos were electron neutrinos, not muon neutrinos, and they had rather different energies (~MeV rather than ~GeV). So it’s a nice indication, but I’m sure theorists could be inventive enough (if necessary) to explain why the effect might be energy dependent or flavour dependent …

  9. Rhodri Evans Says:

    Why don’t they repeat the experiment with neutrinos as similar as possible to those from SN1987A? Is there a technical reason why CERN cannot produce those?

    Also, on my earlier concern about the claimed accuracy of their distance from source to detector of 20cm over some 738km, surely a neutrino detector is many metres in size. Do we know THAT accurately which atom detects a given neutrino?

  10. The fundamental problem with this experiment is that it doesn’t do a like-for-ike comparison: there is no reference light beam that travels the same path as the neutrinos. This means that they have to rely on extremely accurate calibrations of the distances, which I am not convinced about at all. Do they really know the distance as accurately as they claim, given the size of both source and detectors? Is the distance even stable to that accuracy?

    • Yes Peter, that’s the point I’m trying to make. I;m sure they can measure the time to the required accuracy, but I can’t imagine the distnace can me determined to a 20cm accuracy over the 750 odd kms. And, distance from what to what? Neither the source nor the detector is a pont souce.

  11. cenorthChris North Says:

    The weak point would seem to be the translation from GPS position on the surface to the end of the tunnels 140m down. If those tunnels were drilled relative to the “local vertical”, then the Gran Sasso mountains would affect that by a small amount, which may not be insignificant on these scales. If so, this would not so much be a discovery of fundamental physics, but rather an indirect measurement of the local geophysics.

    • I don’t understand the point you’re trying to make here Chris. But, first, a few questions (saves me looking it up). What are the tunnels you refer to? I realise both the source, and of course the detector, must be below the surface, and GPS can only work from the surface, so somehow they have got to go from a distance measured at the surface to the actual distance from source to detector. So I wonder how they do this?

  12. The discussion of their measurement of the baseline and its accuracy starts on page 9 of the paper.

  13. Matthew Bentley Says:

    Assuming the experiment used a properly measured baseline and properly synchronized atomic clocks, then what the experiment actually measured was the local speed of light along that particular baseline. Up until now the speed of light has only been measured at the surface of planet Earth (correct me if I’m wrong). The neutrinos didn’t break the light barrier, they simply showed us that the constant c is not so constant after all. Einstein’s postulate just may have been proven incorrect.

  14. This is a great article on the topic that I’ve just seen –

  15. Anton Garrett Says:

    The height of the top of the Wrekin, Shropshire’s iconic hill, is different by many metres acording to Google Earth and according to the geometric calculations of the Victorians. Google Earth gives 390m, whereas 3D triangulation gives 1335ft (ie, 407m) as quoted on the stone at the top and on The Wrekin’s Wikipedia entry. Bet the
    Victorians are right! Is Google Earth’s figure based on GPS data…?

  16. Dave Carter Says:

    Not sure that we should be using Google Earth as any kind of reference. What does the Ordnance Survey say? If thats different from the stone I would be surprised.

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: