Astronomy Without A Telescope – FTL Neutrinos (Or Not)


The recent news from the Oscillation Project with Emulsion-tRacking Apparatus (OPERA) neutrino experiment, that neutrinos have been clocked travelling faster than light, made the headlines over the last week – and rightly so. There are some very robust infrastructure and measurement devices involved that give the data a certain gravitas.

The researchers had appropriate cause to put their findings up for public scrutiny and peer review – and to their credit have produced a detailed paper on the subject, beyond just the media releases we have seen. Nonetheless, it has been reported that some senior members of the OPERA research team declined to be associated with this paper, considering that it was all a bit preliminary.

After all, the reported results indicate that the neutrinos crossed a distance of 730 kilometres in 60 nanoseconds less time than light would have taken. But given that light would have taken 2.4 million nanoseconds to cross the same distance – there is a lot hanging on such a proportionally tiny difference.

It would have been a different story if the neutrinos had been clocked at 1.5x or 2x light speed, but this is more like 1.0025x light speed. And it would have been no surprise to anyone to have found the neutrinos travelling at 99.99% of light speed, given their association with the Large Hadron Collider. So, confirming that they really are exceeding light speed, but only by a tiny amount, requires supreme confidence in the measuring systems used. And there are reasons to doubt that there are grounds for such confidence.

The distance component of the speed calculation had an error of less than 20 cm out of the 730 kilometres path, or 0.00003% if you like, over the data collection period. That’s not much error, but then the degree to which the neutrinos are claimed to have moved faster than light isn’t that much either.

But the travel time component of the speed calculation is the real question mark here. The release time of neutrinos from the source could only be inferred as arising from a 10.5 microsecond burst of protons from the CERN Super Proton Synchrotron (SPS) – fired at a graphite target, which then releases neutrinos towards OPERA.

The researchers substantially restrained the potential error (i.e. 10.5 microseconds) by comparing the time distributions of SPS proton release and neutrino detection at OPERA over repeated trials, to give a probability density function of the time of emission of the neutrinos. But this is really just a long-winded way of saying they could only estimate the likely travel time, more or less. And the dependence on GPS satellite links to time stamp the release and detection steps represents a further source of potential measurement error.

Some of the complex infrastructure required to infer the travel time of neutrinos across the OPERA experiment. Credit Adam et al.

It’s also important to note that this was not a race. The 730 kilometre straight-line pathway to OPERA is through the Earth’s crust – which is virtually transparent to neutrinos, but opaque to light. The travel time of light is hence inferred from measuring the path distance. So it was never the case that the neutrinos were seen to beat a light beam across the path distance.

The real problem with the OPERA experiment is that the calculated bettering of light speed is a very tiny margin that has been measured over a relatively short path distance. If the experiment could be repeated by firing at a neutrino detector on the Moon say, that longer path distance would deliver more robust and more convincing data – since, if the OPERA effect is real, the neutrinos should very obviously reach the Moon quicker than a light beam could.

Until then, it all seems a bit premature to start throwing out the physics textbooks.

Further reading:
Adam et al Measurement of the neutrino velocity with the OPERA detector in the CNGS beam.

A contrary view – including reports than not all the Gran Sasso team are on board with the FTL neutrino idea.

50 Replies to “Astronomy Without A Telescope – FTL Neutrinos (Or Not)”

  1. an error of less than 20 cm out of the 730 kilometres path, or 0.03% if you like

    No, I don’t like.

    0.03% is 20 cm in 730 metres, not in 730 kilometres.

  2. What is the error margin? 0.03%? 0.003%? 0.0003%?
    Using any of these values we still don’t come to a nice round number of 20….

    1. 0.2 meters / 730,000 meters = 2.74 x 10^(-7) = 274 parts per billion = 0.00003% (rounded to one significant figure)

      1. Steve Nerlich,

        You missed a zero in rounding up, that is all.

        0.00003% or 0.0000274% . Check 🙂

  3. Steve, I think you mean “It would have been a different story if the neutrinos had been clocked at 101.5% or 102% light speed, but this is more like 100.002% light speed.” since the actual number reported is 20 parts per million above light speed.

    1. I meant that if the difference had been proportionally more substantial there would be less of a problem around whether the instrumentation is sensitive enough to claim that the tiny difference detected has significance. But yes, a problem with % – thanks

  4. Andrew Cohen and Sheldon Glashow in arXiv:1109.6562v1 make the observation that superluminal neutrinos would radiate by the same effect as Cerenkov radiation when particles exceed c/n. There would be a field theoretic response similar to a supersonic shock front. That makes it clear, if it wasn’t already, that the claim of superluminal neutrinos was wrong.


      1. The first and second pages in the Cohen-Glashow paper:

        pretty much explain this. The bit about the W-loop, which can also include a virtual Z^0, illustrates a path for the generation of e-e^+ pairs. The weak interaction on a small scale is similar to the EM field, and a particle with the weak charge traveling at v > c has field lines which can’t remain constant in that frame. In the case of the EM field this means the field lines adjust in a way which produces radiation for v > c as Cherenkov radiation. Here of course c = c_0/n, for n > 1 the index of refraction. The weak field will respond similarly. The field will respond by increasing the amount of energy which can be transferred to the environment through the virtual W^{+/-} and Z fields. The real effect will then be to produce e-e^+ pairs, ?-anti-? pairs and gamma ray photons.

        Equations 2 and 3 are derived from the Fermi coupling G_f/( ?c)^3. The interaction is weak, but it still exists. The weak interaction is largely weak because it occurs on a longer time scale ~ 10^{-10}sec. So if there are neutrinos traveling faster than light then given the 730km distance, and c =~ 3×10^5km/s it is not hard to see there is plenty of time for these speedy neutrinos to emit this Cherenkov-like radiation.


      2. wonderful paper. Please add to the list of theories that will need to be revised if the result holds. Physics is exciting again.

      3. Quantum field which propagate on a tachyonic world line are not ordered T(?_1?_2 … ?_n) with ?_i = ?(t_i) so
        T(… ?_i … ?_j..) for t_i < t_j which defines some process with an S-matrix S(?_1?_2… ?_n). Without this there is no compact support in a Stone-von Neumann theorem sense. The main problem is not so much for relativity, but for quantum field theory. In effect there is no causal ordering of spacetime events, which means quantum fields do not propagate in a reasonable manner.


      4. How fast is speed loss exactly? What would have to be starting speed to have the same results? If true this explains supernova non FTL neutrinos.

  5. There was already so much evidence that neutrinos did not travel FTL (SN1A, anyone?) that an experiment showing such a tiny delta over FTL just screamed out “MEASUREMENT ERROR”.

    1. Right, except we (or at least I) can’t live with the cognitive dissonance of speaking as if that is what has been shown as of yet.

    2. SN1a, 1987a and the like are neutrinos travelling through vacuum. OPERA muons travel through earth.

  6. I’m wondering if they could use pulsars to check the timing. Personally I have a wish fulfillment on Einstein being right just as epicycles were right i.e. the model worked until more data became available and it’s not the whole story. The prospect of a light speed limit in a universe so large means we might as well be in a cage. It will be interesting to see if everyone is so keen on Einstein’s rules after we detect that 1.05 Earth size planet with indications of oxygen, nitrogen, methane and chlorophyll orbiting a G class star in the Goldilocks zone and it’s 800 light years away?

    1. Light speed is not really the barrier it appears to be. Time dilation means you could travel anywhere in the universe you like at 99.99999 (etc) % of light speed within your lifetime as long as you don’t want to go back home where billions of years may have passed since you left.

      The problems in achieving this are – a) the acceleration will kill you, and b) colliding with a hydrogen atom at 99.99999 (etc) of light speed will kill you.

      If we can solve these issues, the universe is ours.

      1. Well, it is a gilded cage I would say.

        As you note we can still colonize. We can even have an information exchange market.

        However, other goods would be too expensive to exchange. For all practical purposes what makes relativity speak of local frames of physics (to make universal laws possible) makes it speak of local frames of society.

        Societies will always be local and differentiated, far from Star Wars’ galaxy spanning empires. If you want to wax philosophical, relativity also makes universal laws of biology possible (as in: evolution will rule, markedly).

    2. Even pulsars gradually spin down over time, our clocks are more accurate than that. That’s why they’re part of the plaques and record covers of the solar escaping Pioneer and Voyager probes. If anyone finds it and recognizes the ‘pulsar map’ for what it is, and the binary notations of their pulse rate for what they are, and compare that to the pulsar spin rates at that time, they’ll not only know approximately what location the probe came from, but how *long* it’s been in transit…

  7. it seems to me that speed is speed is speed…..

    the best martial artists in the world know that there is always someone out there bigger or better, and i don’t see why light thinks it’s so special that nothing out there can beat it. sure it’d be nice to have a nice speed limit to go by, but i see it as being unrealistic; i mean, just because someone posts a sign on a freeway saying 55 mph do you think everyone follows it?

    what do you smart people that know what your talking about have to say about that?

    1. ‘the best martial artists in the world’

      If your notional ‘martial artists’ were ‘the best in the world’ then there could not be any other ‘martial artists’ who is bigger and better. If there was a bigger and better martial artist then they would be the best in the world wouldn’t they and not your miserable pretender who thinks he or she is the best.

      We are dealing here with some thing that is well known and studied and the math’s have been run through so many times – so many times – even the device that you used to make your reply here would not work if we didn’t know this math and this wasn’t a indisputable fact of the universe that surrounds us, so instantaneously this result throws up warning lights. The scientists behind this research have looked at their results and asked their fellow scientists to check their results. I see lots of people like you telling smart people they don’t know what we are talking about – go stand in front of the computing device you used to make your post and think how you could have said ‘what do you smart people that know what your talking about have to say about that?’ if that device did not exist – that is how accurately we know this math, which is why we do not jump to conclusions and make rash statements.

      1. ok again, but without the mocking tone this time. i’m not giving you guys that accept the special relativity as fact a hard time. i personally think that humankind is the same dumb beast who drew maps of a flat world and thought we were the center of the universe.

        they went by what they knew at the time, which we are also doing in the present. you say it’s all backed up by math and is 100% accurate, but that isn’t true is it? because everything breaks down when you get into quantum levels or so i’ve heard. so special relativity isn’t finished yet, it doesn’t answer everything, and can’t explain some phenonema. now you seem very smart and you’re very sure of yourself.

        everytime i read about this “speed of light” though…. idk.

        light is a wavelength right? what’s so different about light and sound? are they not barriers that can be broken? what property of physics says that, “um, hey guys, you know, we can’t have things moving around faster than their shadow, cause, that, would, like, cause you to go back in time, and if you don’t believe me, then, check out this equation: T= (t`+l)/ac+1a) and a > 1.”

        ok i’m starting to troll a little bit, and probably sound dumb…….
        obviously einstein is smarter than me and probably understands intuitively what i would need to go to college for 20 years to get a grasp of. so if he says that light is tied to spacetime in ways that prevent anything moving faster than it, it must be true… unless these neutrinos are disobeying traffic laws. geeze i wonder if cosmic tickets are as outrageous as californias, (almost cost 400 dollars for a speeding ticket!!)

        TL DR: why can’t anything travel faster than light?

        (wikipedia isn’t very helpful on this subject, they have all of 1 paragraph about it)

      2. Special Relativity is tested over and over again. And SR and quantum mechanics are fairly well compatible. As a matter of fact, the spin of an electron is a direct effect of a combination of SR and QM.

        The one that is incompatible to QM is General Relativity which includes the effect of a curved spacetime, aka gravity.
        On the other hand, QM and GR are the best theories ever. They are repeatedly tested over and over again, and they pass them all. Your GPS receiver wouldn’t work without GR. That’s just one example where you use it in your every day life.

        So, please. The limit of lightspeed is well tested and confirmed. Don’t form any hoaxes around it, especially if you don’t really know what you’re talking about.

      3. Here we go again. SR is NOT well tested. It says “speed of light is constant for EVERY observer” but have we ever put equipment at the speed of light? NO. It says “time dilates” and we have have observed that from earth for travelling muons, but have we ever observed what the muon say? NO. The theory has not been validated for EVERY observer, just a special class of observers.

        I shine a light beam onto the wall, 2 photons go together at C holding hands. Except one is going at C with respect to the other and vice versa. WFYM?

      4. Why anything can’t travel faster than light: Because according to Einstein’s theory of relativity, which is one of the most successful and well-validated physics theories, tells you that you would need to put an infinite amount of energy into a particle that has mass, if you would want to accelerate it to the speed of light.

        Neutrinos have a very small mass, which means that according to Einstein’s theory it is not possible to accelerate them to the speed of light.

  8. Hi Steve – Another nice article – Just one curiousity… Why say, “..would have taken 2.4 million nanoseconds ..” – Why not 2.4 Microseconds ??

    1. Because 2.4 Million Nanoseconds are 2.4 Milliseconds, or 2400 Microseconds. 😉

      The other reason is that he wanted to compare the 60 Nanoseconds directly to 2.4 Milliseconds. However, with 2.4 Million Nanoseconds it’s a bit more catchy what you are comparing. At least for me it gives a much easier perspective between the two.

  9. Also wondering if Tobjorn’s hypothesis might be correct that the Earth’s geode may be responsible for this discrepancy, ie. gravity could redshift the light and the neutrinos are not prone to such interaction with baryonic matter hence gravitational influence? (I rephrased Tobjorn’s hypothesis and mixed it with some other reader comments to be fair) ?? Thoughts.. Bueller, anyone???

    1. The Earth’s gravity is far too weak to induce this measurable change. Further, redshifting and time dilations in a gravity field are most pronounced for light and particles moving in a radial direction.


    2. Fair enough, but for the record, the physics behind my suggestion was totally different. See my comment above where this is briefly restated.

      Also, you are not the first of UT commenter’s to misspell my name recently, so don’t take it personally that I try to correct it here: it is “Torbjorn”. I realize it is difficult to decode for english speakers. But here is the trick: “Tor” stands for the asa god Thor, a recent movie hero btw. Easy enough to see (and remember) if you note that I am scandinavian.

      [If you want the context of the trick, it is this: “björn” is swedish for bear. So my first name means “Thor’s bear”. Likely a ferocious beast which packs a punch. Right? “Anyone? Anyone?“]

  10. Do you know if they have brought the atomic clocks back together for a comparison? The small difference in gravity between CERN and SG may not be much when you weigh yourself, but it could be enough to give you 60 ns of difference over a week or so. … See this paper by Carlo Contaldi:

  11. I am puzzled by the path the Neutrinos have taken from source to destination. If they have traveled through earth crust between two points on the earth surface, that means midway on that path they have come closest to earth center. The gravitational forces along that path must have changed. It other words, the space-time distortion due to proximity to earth’s center-of-mass must have changed along the neutrino’s path. Now since that was the actual path each neutrino has taken in the measurement through distorted space-time, shouldn’t light speed be measured on the exact distorted space-time line to compare it to? If the comparison was only made to “calculated” speed of light without taking into account the full impact of space-time distortions then wouldn’t that introduce the anomaly? Or was it taken into account? I don’t remember seeing this in the paper. Can someone shed some light on this :)?

    1. It was never a race – they just calculate that the neutrinos travelled faster than light would move in a vacuum over that same straight line distance .

      Relativity requires that light moves in a vacuum at a constant (i.e. light speed) regardless of context – so gravity is not really an issue in the way you describe.

      They didn’t consider gravity effects on neutrino speed – but presumably any effect would be of a slowing nature, so would not be relevant to them seeming to move FTL.

      1. Well, at least they need three more tests to support FTL neutrinos. But this time it should be in different directions – just to discount Earth’s gravitational / rotational effect. Is it possible for them to do the tests from CERN to ISS?

  12. This article, and Butterworth’s are really good! They also seem to touch many of the bases for the physics and the problems with the result (seen as eventual error or eventual valid).

    As for what could predict this result, I don’t think my suggestion of checking for somehow inadvertently measuring along the geode pans out by the way. I looked again, and they do seem to use an implicit conversion to cartesian coordinates. But again, the GPS system and its documentation is wast. And the effect is remarkably corresponding to the discrepancy.

    A better suggestion was offered somewhere else, it was noted that the geode isn’t all that precise. Or at least it didn’t use to be, a revision changed the coordinates a few meters. So these revisions, which should be tracked by the GPS parameter changes and are stated up front in the programming interface, needs to be checked as well.

    But most I like Butterworth’s concern, about the problem of defining the wavefront. (And with it, “information” travel.) This is what analogous light guide “ftl” experiments rely on to show seemingly superluminal effects. So that is very much precedented.

  13. The moon short of test doesn’t work under all explanations of this data that involve new physics. One way of explaining this data is if there is an intermediate particle that is a tachyon. So rather than directly produce neutrinos, the reactions we think produce neutrinos produce a short-lived tachyon that decays into a neutrino. This has the advantages of explaining the OPERA data while also being consistent with the SN 1987a data (since a few nanoseconds of lead time won’t at all matter in a situation where the estimate is at best within minutes). We might have noticed a real result here that is only noticed because the length being measured is fairly small.

    While this is a neat possibility, it still seems much more likely than not that this data is due to some sort of mistake.

    (Disclaimers: I’m not a physicist.)

    1. a tachyon is that you need infinite energie to slow it down to the speed of light. How less energie how faster it go. THis neutrino is different than that. If you use more energie it go faster. So i do not think it is a tachyon. Why would a tachyon move so close to the speed of light ?

      1. Right, that certainly seems to be a serious problem with simple versions of the tachyon claim. There’s no reason to think that the tachyons would be really hih energy. There are some variations of this hypothesis that are being thrown around. See for example John Costella’s post where he suggests that the tachyons are moving nearly instantaneously (from very low energy) but are only stable for the 18.2 meters of the hadron barrier.

    2. I don’t get the first sentence, but the rest seems plausible enough. A tachyon interacting by decay could very well result in non-tachyon particles, the preserved impulse would then be seen as larger massed slower moving particles.

      1. Er, replace “short” with “sort” and my sentence should be clear. The idea is in reference to the article’s comment about using a neutrino beam from the Earth to the moon or some other long distance. My point is that that won’t necessarily be a good test.

  14. Just a (dumb?) idea: is the speed of the earth around the sun (or around the milky way) calculated in? Gran Sasso might have advanced a little bit towards the source, in a cosmic sense.

  15. Just a question: is the speed of the earth around the sun (or around the milky way, for that matter) accounted for in the calculations? Perhaps Gran Sasso moved slightly towards the source, in a cosmic sense.

  16. Just a question: Has the speed of the earth moving around the sun (or around the milky way, for that matter) been accounted for? Perhaps Gran Sasso moved slightly towards the source, in a cosmic sense.

    1. The used GPS shows that relativity still works for EM, so they haven’t been using “aether” to establish distance or timing.

      The neutrino physics is open season however. I assume if you go the extra dimensions route you have to consider such ideas.

  17. I believe that since at least one postulate used to create Einsteins Theory of Special Relativity is not true there is some room for debate here. The Lorentz Transformation was used to postulate nothing can travel faster than the speed of light. It is nothing more than math to describe an optical illusion. No frame in spacetime can be fixed. And if this is all based off of that and relativistic mass then I think everyone should have an open mind and not just beat their chests over how smart they are. Relativistic Velocity should be considered as an alternative.

Comments are closed.