Of Particular Significance

Supernovas and Neutrinos

POSTED BY Matt Strassler

POSTED BY Matt Strassler

ON 09/20/2011

Supernovas are some of nature’s most common and powerful nuclear bombs.  They are also among the most useful for particle physicists and astrophysicists alike.   [The last quarter of this post, on the OPERA experiment’s claim of light traveling slower than neutrinos, has been updated a couple of times, in obvious italicized remarks, in response to some very useful comments.]

In “core-collapse” supernovas,  a huge number of protons are converted, by the absorption of electrons, into neutrons, with the consequent emission of neutrinos.  [Powering this process is one of the most important natural roles of the weak nuclear force.]  Somehow — and active research continues on this topic — the resulting shock waves (perhaps aided by something we don’t know about yet?) blow the star apart.

One of the most exciting events ever to occur in human astronomy was the 1987 explosion of a giant bluish star in the Milky Way galaxy’s largest satellite galaxy, the Large Magellanic Cloud, a luminous blob that is easily visible south of the equator.  Astronomers looking with the naked eye one night in February 1987 spotted a star in the Cloud that they were quite sure shouldn’t have been there.  This simple naked-eye observation began a great wave of astronomical activity that swept round the southern half of the earth, as every astronomer who could do so sought to take advantage of this once-in-a-lifetime opportunity.

A supernova is so bright that its light can briefly outshine the entire galaxy that contains it.  Yet only a small fraction of a supernova’s energy is emitted as light, or in other forms of energy that are eventually converted into light. Most of the energy of a supernova streams out, unseen, in the form of the above-mentioned neutrinos.

Here are some astounding numbers, quoting from a pedagogical website maintained by Steven T. Myers, Tenured Astronomer,  National Radio Astronomy Observatory, Socorro, NM : Almost all of the energy of the [1987] supernova came out in lightweight weakly-interacting neutrinos. About 1058 neutrinos were produced in the core collapse. On Feb 24, 1987, about 1013 neutrinos passed through your body from the supernova! About a million people on the Earth had an “interaction” with a neutrino, of course with no noticeable effect.

That’s right: something like 10 trillion neutrinos passed through your body after the explosion of a star more than 160,000 light years away, several times further than the center of the Milky Way.  Amazing, this universe of ours.

Thousands of trillions of neutrinos passed through several neutrino detectors on earth, and of these a tiny handful — about two dozen — hit something on their way through.  These collisions were recorded over a period of about 13 seconds. No one was paying particular attention at the moment that this occurred, but after the supernova was noticed, experimental physicists went back and found this flurry of neutrino collisions in their data.  The flurry occurred about 20 hours before the first observation of the unexpected star in the Large Magellanic Cloud.  This discovery represented the birth of neutrino astronomy, now an active field of research.

Meanwhile, a look back at older photographs revealed one showing visible light from the supernova that was taken only 3 hours after the neutrinos had arrived on earth. Since the shock wave from the supernova blast had to make its way out of the exploding star before the debris could begin to shine, whereas the neutrinos from the explosion could sail right through the star unimpeded, a delay of a few hours between the arrival of the neutrinos and the arrival of the light was expected.

Now this story is wonderful and fascinating, but why am I mentioning it now? There are two reasons.

First, a relatively nearby supernova has been observed recently, and astronomers are very excited about it.  But press reports don’t seem to have a sense of scale.

  • There have been numerous headlines saying “Brightest Supernova in 40 Years“, “Youngest and Closest in Decades“.  I do not think one needs a degree in physics or astronomy to calculate the time since the great 1987 supernova.
  • The current supernova is over 20 million light years away, more than 100 times further away than the 1987 supernova.
  • The current supernova can be seen by amateurs, but only by those armed with a good telescope, or maybe good binoculars in dark skies; the naked eye is insufficient.  [Its brightness has recently peaked; look soon!] The 1987 supernova was bright enough to be seen easily with the naked eye.

Why all the discrepancies?  Supernovas come in a few different types.  1987 saw a Type II supernova, in which the core of the star collapses and protons convert to neutrons as described above, with the ensuing neutrino blast. The current supernova is a Type Ia, which explodes through a different though not completely understood mechanism.  Type Ia supernovas are very important in astronomy, since they show considerable regularities that can be used to estimate their distance from earth, a fact that played a central role in the discovery that the universe’s cosmological constant, sometimes called “dark energy”, is not zero.  So astronomers are very excited to have an opportunity to study a Type Ia supernova in very great detail and with modern equipment, especially one that was discovered so soon after it exploded.  

To sum up, this is the brightest and closest and most scientifically useful Type Ia supernova in several decades (ignoring one from 1986 that was not so easy to see or study), though it is not nearly as bright or close as the 1987 Type II supernova.

Second, there are rumors on the blogosphere of neutrinos being observed to travel faster than the speed of light. A beam of high-energy neutrinos from the CERN laboratory near Geneva (which also houses the Large Hadron Collider) is rumored to have arrived earlier than expected at the Gran Sasso laboratory in Italy, where a tiny fraction of the neutrinos are observed by the OPERA experiment. Keep in mind there’s no official announcement from OPERA yet, so this is just rumor-mongering at this point, quite possibly in error to a greater or lesser degree.

But in any case, one should treat any such claim, if indeed one is eventually made by OPERA in the near future, with considerable skepticism. This is partly because of the observations made in 1987 following the supernova.

As I mentioned, the neutrinos from the 1987 supernova arrived on earth within 13 seconds of each other, and were followed within at most 3 hours by the light from the supernova, the delay being roughly as expected.   These coincidences are considerable evidence that the neutrinos were traveling neither significantly slower nor faster than light, and that they were all traveling at almost the same speed.  Think about it: these neutrinos traveled for 168,000 years, about 5 trillion seconds, and arrived on earth within about 13 seconds of each other, and within 3 hours (about 10,000 seconds) of the light.  If the neutrinos had been traveling 1 part in a million faster than the speed of light, they would have arrived months before the light; a part in a million slower, and they would have arrived months later. And if they had traveled at different speeds by even one part in a billion, they would have arrived not in a 13 second burst but spread out over hours.

In short, there is evidence from this data that the neutrinos traveled at the speed of light to an extremely good approximation — to perhaps a few parts in a billion.

To measure an effect of a few parts per billion on the speed of neutrinos traveling from CERN to Gran Sasso — a distance of 730 kilometers, which light can travel in 1/400 of a second — would require measuring the travel time of these neutrinos to a small fraction of a nanosecond (a billionth of a second). Measuring anything to better than a nanosecond is tough; coordinating clocks 730 kilometers apart to this level of precision [probably not a good argument by itself – see Lubos Motl’s comment below] would be quite a feat.  [Several readers have already pounced on this paragraph, and they are right to do so; it is the weakest one in the whole post.   I didn’t say this feat was clearly impossible, just difficult.  But I do know this:  picosecond (trillionth of a second) timing measurements are rare in particle physics experiments; for instance typical timing at LHC experiments is at best at the 100 picosecond level, and usually a bit worse.  And I can’t think of a reason why OPERA, given its main task — the OPERA experiment has been designed to perform the most straightforward test of the phenomenon of neutrino oscillations, exploiting the CNGS high-intensity and high-energy beam of muon neutrinos produced at the CERN SPS in Geneva pointing towards the LNGS underground laboratory at Gran Sasso, 730 km away in central Italy — would have had clear cause to expend the money and effort to have picosecond-level timing.  (By the way it’s not just timing but also distance that requires an exceptional measurement.) So I suspect there is something crucial that I am unaware of.  This is the problem with rumors; we do not have enough details to make intelligent and wise comments.  Anyway, thank you readers for forcing me to clarify what I meant here, and let’s now just wait for the curtain to rise on OPERA. NOTE ADDED: After the experiment was announced and detailed, it became clear that the effect was, relatively speaking, an enormous one: 60 nanoseconds!]

There are loopholes in my argument.  Maybe neutrinos traveling through the earth behave differently from those traveling through space.  Maybe, of the three types of neutrinos, one type behaves very differently from the other two Loophole Closed.  [And oops! I had meant to mention a third loophole: Or the effect might depend on the neutrino’s energy, and the neutrinos from the supernova have a lower energy from those studied by OPERA.  Thanks to theorist Marc Sher for his comment  pointing out that this is specifically what was suggested by theorist John Ellis (along with Nicholas Harries, Anselmo Meregaglia, André Rubbia, and Alexander S. Sakharov) back in ’08, in a paper that was probably OPERA’s motivation for the current measurement.]  In these cases, perhaps one could imagine neutrinos traveling faster than the speed of light by a larger amount, one that would be easier for OPERA to observe.  But one should keep the supernova from 1987 in mind when evaluating the plausibility of any claim that high-energy neutrinos traveling 730 kilometers in 0.0024 seconds do so at a speed significantly different from that of light.

By the way, if you happen to read articles about this elsewhere, please don’t be confused by people suggesting that OPERA has a “6-sigma observation” of something new. Such language seems at first to imply a very clear sign of something novel and extraordinary — that the discrepancy observed in the data must be a real effect.   But all it really means is that the effect can’t be a statistical fluke, or due to a known systematic effect. It could still be due to a mistake, or something wrong with the apparatus, rather than a new physical phenomenon.

I should emphasize once again that until the OPERA experiment announces its results, we’re dealing in rumor, and rumors are often wrong.   But it’s no rumor that a star, in its final death aria,  can potentially teach us more than any earthbound measurement.

Share via:


65 Responses

  1. Because of these ingredients contain caffeine, which is another of the
    many forms you can purchase these Pure Green Coffee Extract in various online stores and local pharmacies near you.

  2. Hi professor! The media is sometimes confusing, if they reported that a supernova photon stream came from a galaxy 20 million lightyears away, am I right to conclude that the supernova event happened 20 millions years ago? because it take time for the photons to travel. Regards 🙂

  3. In AWT based on dense aether model the neutrinos aren’t real tachyons, they’re just a solitons of longitudinal waves of vacuum (i.e. gravitational waves) and as such only slightly superluminal (you can imagine them as a Falaco solitons at the water surface, which are always moving slightly faster, than surface ripples) and they do represent the lightweight photinos of SUSY. Photons are supersymmetric bosons, they’re always spreading in lower speed than the light waves and they correspond the Russel’s solitons at the water surface. From this reason I do believe, the OPERA result is substantiated well and it will confirmed later in indepenedent observations.

  4. Assuming the result that neutrino’s travel faster than light is eventually confirmed, there could be a way to explain the phenomenon without losing the cosmic speed limit. The simple step is to take the view that light does not travel at the cosmic speed limit, but fractionally slower.

    Neutrinos, which have a tiny mass, also travel slower than the cosmic speed limit.

    Of course the ~60ns result would then imply that light has a fractionally greater mass than a neutrino. This could be explained by the observation that light (across the whole spectrum) has a wave/particle duality. In effect when the photon manifests as a wave, it travels at the cosmic speed limit, and alternatively when the photon manifests as a particle/anti-particle pair that pair then has mass and consequently does not travel at the cosmic speed limit.

    If we can establish the exact mass of a neutrino then perhaps we could use the ~60ns result to calculate what proportion of time photons spend as a particle/anti-particle pair of a given mass?


    1. You said: “Of course the ~60ns result would then imply that light has a fractionally greater mass than a neutrino. This could be explained by the observation that light (across the whole spectrum) has a wave/particle duality. In effect when the photon manifests as a wave, it travels at the cosmic speed limit, and alternatively when the photon manifests as a particle/anti-particle pair that pair then has mass and consequently does not travel at the cosmic speed limit.”

      This is why in physics we do calculations before we spout words. Your last sentence is simply incorrect, as I calculate in my introductory quantum field theory class every year. No matter how massive the virtual particle-antiparticle pair, the photon remains massless. The mistake in your intuition can be rectified in part by reading http://profmattstrassler.com/articles-and-posts/particle-physics-basics/virtual-particles-what-are-they/

      Moreover, we also look at experiments before we make theories. Experiments show that the photon has a mass no greater than 0.00000000000001 eV/c2, far too small for any effect such as the one you are hoping for.

  5. From a surveyors point of view OPERA public note No. 132 doesn’t do much to address issues of systematic errors.

    1. That precision of 20cm, most contributed by measurements in the Gran Sasso tunnel, looks big enough to need an explanation. Related to that, the precision of horizontal directions of 5cc looks too large for the Leica TS30 instrument. The manufacturer’s value is 0,5″= 1,5 centesimal seconds. When the achieved precision is much worse than expected one naturally suspects a problem.

    2. The co-ordinates of the GPS points outside the portals seem too close to each other for carrying accurate orientation into the tunnel. (Table 1) GPS1 and GPS2 are less than 60m apart and GPS3 and GPS4 are less than 80m.
    Although GPS fixes are of ultimate precision (so far as the surveyor is concerned), the relative precision of close points is not that good- they are poorly correlated. Certainly one can fix the relative position of points 60m apart by terrestrial measurements much more accurately than by GPS. It is really surprising that they used such short orienting rays – and then, fixed them using GPS. The internal legs (200m-600m figure 2) are actually longer- so they were extrapolating their external orientation, which isn’t advisable.

    3. Their GPS points are fixed on ITRF2000, but that brings up the issue of changing baseline length with time. The main Adams article has a stunning pic of a tare in position of one of the control points due to an earthquake- amounting to maybe 7cm-which is a sizeable fraction of the public note 132 stated precision. So in the metrology article, one would expect that to be brought up in some way.

    4. The Gran Sasso tunnel orientation relative to the baseline with CERN is optimal for lateral refraction in the survey to have affected the baseline. That issue is likely to concern any surveyor reading the article- its famous. If a traverse is run near one wall of a tunnel, the line of sight is liable to bending. One way to control for that having a catastrophic effect is to control direction using a gyro instrument- even though that has lower precision. The article mentions that preparations were done for later gyro measurements- but that hasn’t been done yet.

    5. Figure 2 showing the field method, together with reported use of two reflector prisms, raises an issue of orientation drift during measurement. If there were only 2 reflectors then from the middle station, 4 sights to reflectors could only be taken if the reflectors were moved during the measurements. Its hard to see how 4 sets of measurements from each station could then have been done- unless the direction measurements were only those along between stations Then any contribution those measurements across to the other wall of the tunnel could make towards controlling lateral refraction would have been lost.

    6. Everyone can have sympathy with the surveyors who needed to complete this work in less than a week. If one puts together the rush with the loose accuracy requirements they faced (unlike the task confronting the surveyors who controlled the original driving of the 10km tunnel from either end) – then the warning flags go up. They were working in almost an opposite environment also from the physicists who were squeezing the last drop out of their methodology.

  6. Hi Matt. Back to issue of GPS, we know that there are not two GPS measuring absolute numbers at the same size. And we know that the U.S. government prevents absolute measurements. Another issue is that the best performers do not have absolute GPS precision. And finally, we know that the labs are far from open and they gets more difficult to the exact location of each instrument inside them. As those 730 km distance do you believe that there are no measurement error? Thanks

    1. I agree this is something to watch, but I know I am not expert enough to have an useful opinion. On this point one must rely on an array of experts in metrology reviewing the subtleties. I suspect a panel is already being assembled.

  7. Dear Matt, I want to share a thought from the outside, and here I’m making some assumptions:
    1) The results will be confirmed by further experiments (a big if).
    2) The theory of general relativity still holds, at least in this corollary: superluminal velocities mean information flowing from the future to the past.
    3) The information sent from the future can influence the past.
    Doubts about the first. Almost a century of passed tests for the second, but not for the corollary. Assume the third since there is information flow by “faith” in point 2 and the universe making non quantum information inusable would be even stranger IMHO.
    From a computer scientist point of view this would mean that NP-complete problems would be easily tractable. Despite not being proved to be outside the P class (roughly speaking: the tractable problems), they are widely believed to be. This would be bigger than breaking the speed of light.
    Here there is a beautiful paper suggesting a connection between computer science and physics: http://www.scottaaronson.com/papers/npcomplete.pdf
    Any thoughts?

  8. I follow along this type of thing as a hobby, so these might be stupid questions … I am curious though. There is a notion here that the types of neutrinos detected from the 1987A Supernova and those used in the OPERA experiment are different. Is this coupled with a notion that 1987A would never have produced the same particular type/energy (or even greater energy) of neutrino as produced for OPERA? Or, if 1987A would have produce OPERA (or greater energy) like neutrinos, that they would not have been detected?

    1. The differences are mentioned in this very post: look at the third-to-last paragraph, the one about “loopholes”. Yes: the neutrinos from the supernova are of mostly created from electron-proton conversion to neutron plus neutrino, and are therefore electron-neutrinos initially, while those at OPERA are created in pion decays to antimuon and neutrino, and so those neutrinos are muon-neutrinos initially. The supernova neutrinos have energies of about 0.01 to 0.05 GeV, while those at the OPERA experiment have energies of around 10 — 30 GeV (numbers approximate.) And the supernova neutrinos traveled first through the star, then through lots of empty space, then perhaps through a good fraction of the earth, while the OPERA neutrinos traveled through 730 kilometers of earth. Any one of these differences might or might not be important, at least in principle, in interpreting the difference in the two results.

  9. Hi Prof Matt,
    It is been really long time since I opened a book on Quantum Mechanics, but this latest news on FTL neutrinos, my interest in the field got a little spark.
    I have a rather newbie-ish query – once a neutrino is formed and travelling at certain speed – is it even possible to accelerate it to higher speed and how, theoretically speaking. If yes, has it been verified experimentally?

    I ask this, as with my superficial understanding, neutrinos are electrically neutral and weakly interacting. So, it must be extremely difficult to ‘energize’ them to higher state.


    1. First a clarification — let’s remember what it means to accelerate an electron. We can do this with an electric field (pointing opposite to the direction of acceleration, because of the electron’s negative charge) or with a gravitational field (consider an electron falling into a neutron star or black hole, which sport large gravitational fields much larger than at the sun’s surface.) If the electron is moving much slower than the speed of light, then our ordinary intuition holds: pushing an electron makes it go faster and makes its motion-energy increase. If the electron is moving close to the speed of light, however, pushing an electron does not make it go much faster, even though it makes its motion-energy increase. In short, “acceleration” for particles traveling near the speed of light results in an increase in energy but only a tiny increase in speed. This is how the universal speed limit is obeyed, in Einstein’s theory of relativity.

      Now, what is different for neutrinos? We can’t accelerate them using electric fields, but we could in principle do so in a gravitational field; neutrinos would accelerate as they fall into a neutron star or black hole. So theoretically there is no problem. But while it is easy to design a powerful electric field to accelerate electrons, gravitational fields of the size required to see a significant effect on neutrinos are out of reach.

      So to verify Einstein’s results by taking neutrinos and accelerating them experimentally would be very difficult. In short, to see them break the light-speed barrier through direct experimental measurement would be difficult to do directly. But it is easy to make a beam that has neutrinos at many different energies (in fact it is hard to make a beam that has a narrow range of neutrino energies!) So rather than accelerate individual neutrinos, what will be done is to measure the speed of neutrinos that have different energies at the time they are produced (in the decays of pions that were themselves produced by collision of protons with ordinary matter.)

      p.s. Accelerating something is not the same as putting it in a “higher state”. When you accelerate something you increase its motion-energy. When you put something in a “higher state”, you are increasing its internal energy. I can put a hydrogen atom in a higher (or “excited”) state while keeping it motionless; I can accelerate a hydrogen atom while keeping it in its lowest (or “ground”) state.

      1. Hello Sir,
        Thanks a lot for the detailed reply. It was much more than I was expecting and very clearly expressed. Even though you must have answered such queries numerous times already :).

        My rationale behind asking ‘accelerating neutrinos’ was – in the remote likelihood of OPERA’s FTL neutrinos proven to be true – I guess it would be relatively safer to assume that those neutrinos were travelling at FTL since their birth and not accidentaly or intentionally ‘accelerated’ within the setup breaching the ‘c’ threshold in the process.
        Of course, there have to be many more possible hypothesis, but making them accelerate in the setup from subluminal to FTL would be least likely of all.


  10. An another thing the earths crust is constantly changing and flexing as the earth spins making the distance between the two measuring points constantly changing not noticable at normal scales but would this need to be conpensated for at the experiments scale?

    I don’t doubt that I don’t understand the complexity of the physics involved (I haven’t studied physics at all but it’s something I am seriously considering) but it would seem that the complexies of measuring such small and fast moving objects over such a realativley small and everchanging distance are enormous and any errors at this scale would be exagerated, something that the team have no doubt accounted for.

  11. I just seems to me that the scales that these guys are working to any microscopic deviation in the gps measurement would distort the result and the gps system is in a constant state of flux as the satellites will never be stationary with each other or the signals synched with each other precisely enough (long enough for pico scale events) to give the accuracy required for this type of measurement. I would of thought that the gps signal just travelling through the air a hitting different densities of ever changing particulate matter in the atmosphere alone (without considering minute changes travelling through buildings etc I don’t know how the detection is done) would provide enough variation in the frequency of the timing pulse at these scales to produce a noticeable effect with the drift, or that microscopic changes in the gravitational fields would affect the timing. Due to the scales involved I would of thought to compensate for the drift and changes in gravitation field strength they would need to be reverifying the distance measured and compensating at the same rate at which they are measuring the rest of the experiment to get a correct result. But like I said it’s just a thought. I guess as a layman that when I just think of the inherant inaccuracies that are systemic in the gps system I can’t imagine how you could reliably get the accuracy needed at the scales required.

    Maybe these guys need to build a very long straight vaccum tunnel to measure the distance directly through the earth admittedly no mean feat on such a scale.

    Anyway no doubt these guys are much smarter than me and have factored in everything.

    1. All these issues can be important in metrology, but they definitely factored all these things in. Of course the individuals involved are smart, but they have more than that going for them: there were quite a few people involved on the physics side, and I’m sure they sought out experienced people on the metrology side. They used the GPS system in a mode which is more accurate than the standard mode. If it were literally impossible to achieve this level of accuracy and precision, they collectively would know that. So the mistake, if there is one, is likely to be something other than what you mentioned.

      A straight tunnel 730 kilometers long and many kilometers deep is impossible. A version of the experiment built over a kilometer or two, however, would allow this.

  12. Just a thought but it says that the team used gps and calculated a statistical result, since the gps satelites are in orbit and in a lower gravitational field where time has been shown to run slower (which causes the satelites signal to drift requiring a daily recalibration) could the distance being measured be varying slightly depending upon at which point in this drift/reset cycle the measuremenrs are being taken. I don’t know how often the team would have reverified the distance and/or how accurate this would be with the drift. Could this slowdown in time and resultant drift cause the variation in the result. I don’t even begin to understand the physics here and I would assume that the team would of corrected for any variations due to this effect. I haven’t read up on the experiment so this may not be relevant the again I not expert on the gps sats like I said it’s just a thought 🙂

    1. Metrologists (such as those who helped these scientists in making these measurements) have to account for all sorts of subtle effects. They certainly are aware of these things. So I doubt it. Of course all of this will need to be independently verified. Even world-class people can make world-class mistakes.

  13. I listened to the presentation this morning (US Pacific Time) and the first part of the question period. I was struck most by the issue of the geodesy. Yes, it’s very good for the geodesy world, but I agreed with the questions about the details of the survey linkage from the surface to the site of the experiment AND the fact that there really are two separate absolute measurements. One is of the time of flight of the neutrinos, which does seem very well calibrated & checked (though who knows what other systematic issues will be found by others). The second is an absolute geodetical measurement of the absolute distance between the two sites — to < 1 m accuracy. But the latter measurement seems impossible to calibrate. If one chose to just ASSUME that the neutrinos traveled at 'c' then the most accurate measurement of the distance between the sites would in fact be the time of flight of the neutrinos. Which begs the question — just how accurate can the completely separate measurement of that distance actually be. Anyway, it will be interesting to see what cross checks and completely separate experiments are suggested and performed. Great to see/hear new science in real-time.

  14. Note that one subtle thing about neutrinos from SN 1987A is that they get slowed down due to gravitational potential of intervening matter along line of sight (aka Shapiro delay). This delay is about 5 months. so strictly speaking, all one can say from SN 1987A is that both light and neutrinos travel on the same null geodesic.

    1. Thank you for this observation (in other words, for the non-experts, one has to remember that in Einstein’s version of gravity, space and time are viewed as curved by the presence of nearby matter, and this means that both light and neutrinos take a path which is not altogether straight.) But I don’t think this much affects the limit as far as light speed versus neutrino speed; any complication from the subtlety would tend to separate the neutrinos’ arrival time from that of the photons.

  15. Using SN1987a as the basis for a conclusive ruling on the speed of neutrinos requires as a prerequisite certainty in our knowledge of stellar interiors. Anyone who has studied stellar structure knows that any such certainty is profoundly misplaced.

  16. Matt, could you please elaborate on your statement “Tachyons are instabilities, not particles”? I am not a physicist, so I ask this standing in a deep well of ignorance. However, when I look up tachyon on Wikipedia, the first sentence I find there is “A tachyon is a hypothetical subatomic particle that moves faster than light.”

    1. I really should, shouldn’t I… I need to write my long-promised article on “Particles, Waves and Fields”, at which point it will become possible to do this. Keep an eye out for that article and berate me if I fail to give an answer to your question.

      As for Wikipedia — well, it isn’t always right, especially when it comes to subtle points in theoretical physics. I have no idea who wrote the Wikipedia article, but part of the purpose of my website is to give people information that comes from experts who are fully engaged in modern theoretical physics, and not locked in to 50-year-old misconceptions. You may as well learn the brand new misconceptions.

    1. OPERA is not part of CERN; CERN just provides the neutrinos. The most likely explanation is that there’s an error somewhere. The next most likely (and it is a *lot* less likely) is that they’ve discovered that the Lorentz-invariance symmetry on which Einstein’s relativity is based is slightly violated. That possibility has certainly been considered before by some very good theorists.

  17. Maybe relativity only applies to particles with Yukawa couplings to the Higgs field. If neutrinos are only left-handed, they can’t have a Higgs coupling, so they don’t have any kind of standard “mass” in the sense that other particles do. So they are exempt from relativity !

    What say you ?

    1. Doubt it. Electrons and neutrinos are converted one to the other by the weak nuclear force; I don’t see how one could be subject to relativity and the other not. But I have no better suggestions at the moment, because I don’t know enough. There are experts who’ve thought about all of the different possibilities, and I have a lot more reading to do.

      1. I don’t mean that they are exempt from relativity in that way.
        Fermions gain their inertia from the Higgs field: this increase of inertia as the speed of light is approached prevents them from surpassing the speed of light. But neutrinos would not feel this same sense of inertia as the other fermions if they did not interact with the Higgs field. Therefore, neutrinos could have a different behavior than electrons in terms of the relationship of energy, mass, and velocity.

    1. Tachyons are instabilities, not particles; neutrinos are not tachyons. However, it is possible in principle for Lorentz invariance (the geometrical symmetry on which Einstein’s theory of special relativity is founded) to be violated, and then the speed of light may not be the speed limit for all particles. There is a serious and detailed theoretical treatment of this possibility, as alluded to in the previous comment by Tim Barrows. Unfortunately I have not studied this issue carefully and have some things to learn before I can comment intelligently. Readers who know more are welcome to explain the situation.

      1. I agree. Tachyonic neutrinos does not form unitary reps of the Poincare group, there must be some sort of Lorentz violation involved

  18. the paper by Ellis et al. mentioned above is a particular case of a more general theory postulated in the 90s by the Indiana group studying Lorentz Invariance Violations (hep-ph/9501341, hep-ph/9703464, hep-ph/9809521). These fellas wrote an effective theory that leads to the modified dispersion relations rather than a phenomenological modification of the speed of neutrinos.

  19. Matt – Part of the systematic analysis in this kind of experiment is being
    sure the events are associated with the beam; the tighter the timing, the
    better. The flavor structure also depends on the timing, since the beam energy is high enough to produce heavy quarks, which can have tau decay modes. Those will be concentrated at the target, while the muon neutrinos whose oscillations constitute the signal will be distributed through the decay volume. That is probably the determining factor; although ruling out overlapping beam-background events, and reducing the emulsion processing are important, precision timing is probably not
    as crucial. Another difference with this experiment is that the events
    of interest are very localized – a tau track is microns long, so the time
    and place of interaction are as well-characterized as you could ask for.
    It allows not only reducing systematic errors, but knowing what they
    are – and for most experiments, that’s most of the battle.

    I could go on, but I wouldn’t know where to stop. Hope this helps; it is still just my impression from looking over one of their recent (on-shell) documents (yesterday), but better confabulated – YMMV.
    Regards, Ryan

  20. My 2c worth: the timing is necessary because Opera is an emulsion
    experiment with a live trigger; to do the short-distance tracking, the
    events are located by energy deposited in active layers between emulsion
    packages, and then emulsion packages are removed for scanning. Having
    a tight timing constraint with the production process reduces the overhead
    in scanning the emulsion, as well as improving signal-to-noise and clarifying systematic effects.

    There are a couple of things that complicate the timing. The proton beam
    arrives at the target in a pulse that has a finite duration, as well as
    (usually) some bunchy structure. That can be measured, even on
    a pulse-by-pulse basis, so the main complication is that not all neutrinos
    have the same start time. Also, the neutrinos are produced in decays
    by a variety of particles (mostly pions), which have different lifetimes
    and different Lorentz boost factors. This aspect requires more modelling.

    The vapor-rumor had the size of the effect at a few nanoseconds, long
    on the scale of proton bunches, but short on the scale of decay lengths
    (gamma-v-tau). It will be interesting to see the analysis when (if) it

    Ryan Rohm

    1. Ryan — thanks for the message… but I had some trouble following some of your points. Is the following correct?

      Pulses of proton beams created back at CERN smash into a target at CERN and produce pulses of pions, which in turn, in their decays, produce pulses of neutrinos heading toward Gran Sasso. A small fraction of these neutrinos interact in the OPERA detector in Gran Sasso, and thereby produce one or more particles that leave a track in an emulsion material. If I understand correctly, these particles are also simultaneously observed by other detectors that have precise timing, which then effectively tells the experimenters: “At this precise time, an interaction of a neutrino occurred HERE.” But I still don’t understand why they need to know the timing very well.

  21. Matt–In John Ellis’ paper, arXiv:0805.0253, he argues that the effect might vary as the square of the energy of the neutrinos, and CERN’s neutrinos are much more energetic than SN1987a’s. He claims that OPERA can, in that case, get stronger bounds than SN1987a.

    1. Thanks for the paper trail! That’s very helpful, and I will follow up. And oops!! I had meant to mention the different energies in my “loophole” paragraph, and forgot to do so.

  22. Dear Matt, I don’t believe that you actually believe that “to coordinate clocks 732 apart with a nanosecond accuracy is quite a feat”.

    Professional clocks in the world may surely be easily synchronized with better than a picosecond accuracy and if you use atomic clocks and electromagnetic signals sent in between them (to measure the exact separation in light seconds, which you want to divide by 2 to get the right delay), you may easily synchronize them with the full accuracy available to atomic clocks which may go down to 10^{-16} seconds or so.

    This can’t be the problem. What may be hard is to figure out when exactly a neutrino departed and when it arrived, relatively to the complex reactions in their vicinity.

Leave a Reply


Buy The Book

A decay of a Higgs boson, as reconstructed by the CMS experiment at the LHC


The particle physics community is mourning the passing of Peter Higgs, the influential theoretical physicist and 2013 Nobel Prize laureate. Higgs actually wrote very few

POSTED BY Matt Strassler

POSTED BY Matt Strassler

ON 04/12/2024

I recently pointed out that there are unfamiliar types of standing waves that violate the rules of the standing waves that we most often encounter

POSTED BY Matt Strassler

POSTED BY Matt Strassler

ON 03/25/2024