Whatever Happened to Ning Li?

[ben m]

Neutrinos travel at the speed of light

Nonsense. The three neutrino masses are known to be in the ranges:

smallest = 0--2 eV
middle = 0.008 --2 eV
largest = 0.05 -- 2 eV

The conventional labels m1,m2,m3 may be in either or two orders: either 1,2,3 or 3,1,2.

 

[WhatRoughBeast]

Yes they do. Neutrinos from Supernova 1987A arrive along with the photons, they don't lead them or lag them. And the recent OPERA results didn't show the neutrinos to be lagging c, which they'd need to do if the neutrinos had mass.

If neutrinos do travel at c, they are "frozen" in terms of their variety. But the answer to the Solar Neutrino Problem is neutrino oscillation. And for neutrinos to oscillate they must be travelling at less than c.

 

[Ziggurat]

Yes they do. Neutrinos from Supernova 1987A arrive along with the photons, they don't lead them or lag them. And the recent OPERA results didn't show the neutrinos to be lagging c, which they'd need to do if the neutrinos had mass.

No. They would need to be sufficiently slower than c to be able to directly detect the difference in speed. But since they have mass (as pointed out above), we know that they do not travel at c. They travel very close to c, but not at c.

 

[ben m]

Also, the "neutrinos arrive along with the photons" statement is also wrong. There's a lot of supernova-physics that *delays* the photons---by of order several hours---and there's no photon-detection event that labels the "moment of photon arrival" from Supernova 1987A. (The supernova light arrived some time before 23:00 UTC, when the night shift at Las Campanas Observatory happened to notice it and sound the alarm.)

Rather, the neutrino "speed" measurement is in fact a *dispersion* measurement. The higher-energy neutrinos and the lower-energy neutrinos, presumed to be emitted within ~10s of each other, arrived within a ~10s window. If the neutrino mass were as large as 15 eV, the lower-energy neutrinos would have arrived later than the high energy neutrinos. There is no such precision to be found by comparing the neutrino arrival time (+/- a few seconds) to the photon arrival time (+/- a few hours).

 

[Cuddles]

Yes they do. Neutrinos from Supernova 1987A arrive along with the photons, they don't lead them or lag them.

Fail. Neutrinos from supernovae arrive before light, because they pass straight through the star when the reaction starts, whereas light is delayed until the shockwave reaches the surface of the star. That gives them enough of a head start that they arrive first even though the light is going (very slightly) faster.

And the recent OPERA results didn't show the neutrinos to be lagging c, which they'd need to do if the neutrinos had mass.

As others have pointed out, neutrinos travel sufficiently close to c that it is difficult to measure the difference. Virtually impossible over any straight line distance possible on the Earth. Even in circular accelerators using electrons which are far more massive than neutrinos, we still treat the electrons as traveling exactly at c because it simplifies calculations and the difference is still so small as to be negligible. But electrons, like neutrinos, have mass, and therefore we know that they're never actually traveling at c, just close enough that we don't always have to care about the difference for practical purposes.

Since experiments for a long time as well as theory predicted that neutrinos were massless, only to be likely shown wrong in the 1990s

The important thing here is that theory and experiments didn't say that neutrinos couldn't have mass, just that they probably didn't. However, both theory and experiments say that neutrinos absolutely cannot have charge. Lacking charge is such a fundamental characteristic that they were even named after it.

Charge conservation is not violated because different neutrinos could have different charges.

This has already been explained to you, so I really don't know why you're repeating it. Charge has to be conserved locally. Look at this equation again:
neutron --> proton + electron + electron-antineutrino
That reaction must conserve energy, momentum and charge, among other things. It doesn't make any difference at all what the charge on different types of neutrino might be, if the charge on the neutrino in that particular reaction is not exactly zero, it's not balanced. And, again as already explained, you can't just add in an extra different neutrino, or any other particle, to try and balance it, because energy and momentum are already balanced so adding anything extra would mess them up and just make things worse.

And note that this is not just theoretical, reactions like this have been constantly measured since before neutrinos were first discovered. In fact, these observations are exactly the reason neutrinos were discovered in the first place.

it's reasonable to suspect that the electric charge will be detected eventually

No, it really isn't. There is no reason to expect it from theory, and there is no experimental evidence to suggest it could be possible. That we have been wrong about things in the past does not make it reasonable to expect every nonsensical claim to eventually come true.

 

[Farsight]

If neutrinos do travel at c, they are "frozen" in terms of their variety.

That's not necessarily true. Take a look at the classical analogy on wikipedia, and imagine the coil spring is a helical wavefunction. Then imagine the whole thing is moving linearly at c. Then imagine it's also "concertina-ing" as it does so.

But the answer to the Solar Neutrino Problem is neutrino oscillation. And for neutrinos to oscillate they must be travelling at less than c.

That's the idea, but the OPERA guys were surprised when the clocked them going faster than c, and decided they'd better release their paper. It turned out that it wasn't a good idea and there's been a couple of resignations, but what's been missed in the hoo-hah is that the neutrinos did not appear to be travelling at less than c.

Cuddles/ben: we don't detect light from a supernova followed by neutrinos hours/days/weeks/months/years later, depending on the distance. Or vice-versa.

Zig: it's the other way round. If they're travelling at less than c, then they have mass. When you trap a massless photon in a mirror-box it adds mass to the system. It's still moving at c round and round inside the box, but it isn't travelling any more. Open the box and it's a radiating body losing mass just like Einstein's 1905 paper. There's a sliding scale to this. If you could make a photon go a little less than c for a spell, it would have a little mass while it did so.

Astrodude: there's some interesting issues with neutrinos, but remember that whilst you can't have charge without mass, the converse is not true. You can have mass without charge.

 

[edd]

That's the idea, but the OPERA guys were surprised when the clocked them going faster than c, and decided they'd better release their paper. It turned out that it wasn't a good idea and there's been a couple of resignations, but what's been missed in the hoo-hah is that the neutrinos did not appear to be travelling at less than c.

That's exactly what you'd expect if they were travelling at the expected less-than-c speed. Look, take the very heaviest mass ben m suggested - 2eV - and work out what speed a particle of that mass would go at with an energy of 17GeV. Then work out the travel time difference over the 730km or 2.4ms path.

Then tell us if OPERA had the necessary accuracy to spot that.

Similar arguments apply to supernovae observations.

 

[Farsight]

Sure, OPERA don't have the accuracy, but they timed them, and it wasn't for nothing. As for supernova, I stand corrected on the "no lead or lag". But if neutrinos are going slower than photons you'd expect to see a distance-related variation in the arrival of the neutrino flash and the primary photons.

Take a look at http://arxiv.org/abs/0712.1750 and read this bit on page 3:

"Just as light passing through matter slows down, which is equivalent to the photon gaining a small effective mass, so neutrinos passing through matter also result in the neutrinos slowing down and gaining a small effective mass".

 

[edd]

A supernova's a complex and messy event, lasting quite a long time. Optically the light doesn't kick in for quite some time, and I'm not sure anyone's caught the initial UV flash you might expect.

Besides, let's repeat the calculation you probably haven't done. Take a 2eV mass neutrino with say 10MeV energy. Take the distance to be 51.4 kpc. Calculate the arrival time difference between the particle and a photon.

Now consider that arrival time error and the difficulties involved when it comes to identifying the supernova in a long exposure image and in a neutrino observatory where you get ten neutrinos arriving, complicated by the astrophysics of a supernova in a star with a radius of approximately one light minute.

Do you really still think this can distinguish between a particle with a mass as above and something travelling at exactly the speed of light?

 

[Cuddles]

Sure, OPERA don't have the accuracy, but they timed them, and it wasn't for nothing.

Correct, it wasn't for nothing. It firmly established that neutrinos travel very close to the speed of light.

As for supernova, I stand corrected on the "no lead or lag". But if neutrinos are going slower than photons you'd expect to see a distance-related variation in the arrival of the neutrino flash and the primary photons.

For identical supernovae involving stars of exactly the same size and composition, yes. But firstly, real stars aren't identical so there's going to be a lot of variation in how much of a head start the neutrinos actually get. And secondly, just how often do you think we even get to measure that? We already have a network of neutrino detectors set up specifically to give early warning about supernovae (imaginatively named the Supernova Early Warning System). How many has it detected since it was first set up a decade ago? None. How many are expected? About 3 per century.

So sure, maybe one day we will be able to measure exactly how much light can catch up to neutrinos from a supernova. So far we have exactly one, entirely accidental, point of data.

 

[ctamblyn]

Besides, let's repeat the calculation you probably haven't done. Take a 2eV mass neutrino with say 10MeV energy. Take the distance to be 51.4 kpc. Calculate the arrival time difference between the particle and a photon.

Since Farsight probably won't, I thought I'd have a shot. For the speed of the neutrinos with mass m and energy E, we've got (in natural units):

sqrt(1 - (m/E)²) = 1 - (1/2)(m/E)² to leading order in (m/E).

So, if L is the distance to the supernova and the neutrinos and photons started off together, the time delay between the photons and neutrinos arriving at Earth is (L/2)(m/E)², for which I get about 0.1 seconds. I.e., not very much.

 

[ben m]

Sure, OPERA don't have the accuracy, but they timed them, and it wasn't for nothing.

Yes it was. It's one of many "let's carry out this analysis of our existing data, even though we have multiple strong lines of reasoning that predict a null result, and no non-null predictions whatsoever" data analyses. All experiments, especially in the neutrino/dark-matter/rare-decays field, do this sort of thing. I've done them myself.

As for supernova, I stand corrected on the "no lead or lag". But if neutrinos are going slower than photons you'd expect to see a distance-related variation in the arrival of the neutrino flash and the primary photons.

Go ahead and make a plot: neutrino/photon time delay on the y axis, supernova distance on the x axis. Now plot all of the supernovae for which neutrinos have been detected. That's ... one point. 1987a. Good luck drawing that line through that point. Also, 1987a is likely to be the *farthest* supernova that we get neutrinos from. It was in the LMC, 50 kpc from Earth. Over the next 100 years you can expect maybe one or two data points in the 0-10 kpc range (Milky Way), over 500 years you have a shot at another LMC or SMC event (50-70 kpc), but events in Andromeda (780 kpc) are borderline undetectable in neutrinos.

You don't seem to understand experimental error. You may expect there to *be* a distance-related variation. That's different than your being able to *see* a distance-related variation.

 

[Astrodude]

Could you explain in more detail how this would work? In particular, how does this account for charge conservation in the beta decay process discussed above?

It couldn't work in the strict beta-decay relationship described above. I was supposing that the system wasn't closed specifically to particles involved in the beta-decay process. If that system was just approximately closed, I was supposing that the charge could be conserved elsewhere, since it would be so small. Based on what I've learned about neutrinos over the past couple of days, I'd say this isn't the case. If neutrinos were to carry an electric charge, there'd have to be another local particle to cancel the charge out, but there doesn't seem to be such a particle.

Still, I'd be curious to see attempts at inducing a charge on a neutrino.

http://arxiv.org/pdf/hep-ph/9305308v2.pdf

The reason that a neutrino has mass, but no electric charge is a problem that a unified field theory of gravity and electromagnetism would have to work out, but it's hardly reason to doubt the existence of gravity-electromagnetism unification in nature. The fact that all atomic-level matter and beyond has mass and electric charge(at least a little in some places) is reason enough to suspect unification in nature.

 

[WhatRoughBeast]

That's not necessarily true. Take a look at the classical analogy on wikipedia, and imagine the coil spring is a helical wavefunction. Then imagine the whole thing is moving linearly at c. Then imagine it's also "concertina-ing" as it does so.

Just exactly how is it going to concertina with a tau of zero? I can imagine a wave function concertinaing at c, but I can also imagine a unicorn, too.

And please give an example of a helical wave function, particularly as pertains to a particle such as a neutrino.

 

[Ziggurat]

The reason that a neutrino has mass, but no electric charge is a problem that a unified field theory of gravity and electromagnetism would have to work out, but it's hardly reason to doubt the existence of gravity-electromagnetism unification in nature. The fact that all atomic-level matter and beyond has mass and electric charge(at least a little in some places) is reason enough to suspect unification in nature.

There's a fatal flaw in the logic of this last sentence. Particles which interact via gravity but not electromagnetism cannot form atomic matter, because gravity is too weak. So the fact that we don't see any in atomic matter isn't because it's not out there in large amounts, but because it won't clump up and we can't see it with light. But the evidence is that there's actually far more of such neutral matter in the universe than charged matter.

 

[Reality Check]

Cuddles/ben: we don't detect light from a supernova followed by neutrinos hours/days/weeks/months/years later, depending on the distance. Or vice-versa.

Farsight: You are wrong. In the only and only (so far) observation of neutrinos from a supernova SN 1987A, the neutrinos arrived about three hours before the light. So we have detected light from a supernova followed by neutrinos hours/days/weeks/months/years later, depending on the distance. Actually vice-versa.

We do not see a distance related lag for the simple reason that we only have one measurement!

 

[Reality Check]

Take a look at http://arxiv.org/abs/0712.1750 and read this bit on page 3:

This is the Mikheyev–Smirnov–Wolfenstein effect:

The precise mechanism for “solar neutrino oscillations” proposed by Mikheyev, Smirnov and Wolfenstein involved the resonant enhancement of neutrino oscillations due to matter effects. Just as light passing through matter slows down, which is equivalent to the photon gaining a small effective mass, so neutrinos passing through matter also result in the
neutrinos slowing down and gaining a small effective mass. The effective neutrino mass is largest when the matter density is highest, which in the case of solar neutrinos is in the core of the Sun. In particular electron neutrinos generated in the core of the Sun will be subject to such matter effects. It turns out that neutrino oscillations, which would be present in the vacuum due to neutrino mass and mixing, will exhibit strong resonant effects in the presence of matter as the effective mass of the neutrinos varies along the path length of the neutrinos. This can result in a resonant enhancement of solar neutrino oscillations known as the MSW effect

I emphasized the bit your quote mining left out. This is a combination of

1. the neutrino masses in vacuum
2. the slowing down of neutrinos in matter giving an additional effective mass.

The MSW effect dominates for high energy neutrinos and these experiments detect the resonances.
Vacuum oscillations dominate for low energy neutrinos and these experiments see detect oscillations.

 

[Farsight]

A supernova's a complex and messy event, lasting quite a long time. Optically the light doesn't kick in for quite some time, and I'm not sure anyone's caught the initial UV flash you might expect.

I said neutrino flash. And that I was wrong about the lead/lag thing. Sorry about that.

Besides, let's repeat the calculation you probably haven't done. Take a 2eV mass neutrino with say 10MeV energy. Take the distance to be 51.4 kpc. Calculate the arrival time difference between the particle and a photon.

I did one, but I lost my post. There's some key-combination shortcut that wipes out everything you've keyed in, maybe a Windows7 thing. I said Pauli predicted the neutrino to account for the missing energy in beta decay, which is typically about the same as the electron mass-energy of 511keV. For a particle with a 2ev rest mass to take away 511keV of energy the gamma factor is about 250,000. So if √(1-v²/c²) = 1/250,000 then 1-v²/c² = 1/250,000² so v is less than c by a factor of 1/62500000000. It isn't much. For SN 1987A which is 164,000 light years away, then with 32 million seconds in a year it's only about 8 seconds. Somebody check my arithmetic.

Now consider that arrival time error and the difficulties involved when it comes to identifying the supernova in a long exposure image and in a neutrino observatory where you get ten neutrinos arriving, complicated by the astrophysics of a supernova in a star with a radius of approximately one light minute. Do you really still think this can distinguish between a particle with a mass as above and something travelling at exactly the speed of light?

No, this is why I said neutrinos travel at c. You can't distinguish the difference. And the important point is that you can't make a neutrino not travel at c in any way you can measure. They aren't like electrons. You just can't have a neutrino just sitting there in front of you. Another important point is the nature of mass. A photon adds mass to a system when it isn't travelling, because it's say going round and round in a mirror box. Then all the photon energy/momentum is exhibited as mass. When you open the box you've got a radiating body losing mass. Now the photon is travelling at c and none of the energy-momentum is exhibited as mass. In between these two limits there's a sliding scale. If you could make the photon travel at less than c then some of the energy-momentum is exhibited as mass. And if photons had some kind of internal dynamics that made them "concertina" through space such that their speed varied, their mass would vary too. As far as we know they don't, but neutrinos do.

 

[Farsight]

Correct, it wasn't for nothing. It firmly established that neutrinos travel very close to the speed of light.

But the media frenzy about the apparent faster-than-light neutrinos missed the trick: the speed of neutrinos would then be c, photons would have an effective mass, and the classification of neutrinos into the same group as electrons falls down.

For identical supernovae involving stars of exactly the same size and composition, yes. But firstly, real stars aren't identical so there's going to be a lot of variation in how much of a head start the neutrinos actually get. And secondly, just how often do you think we even get to measure that? We already have a network of neutrino detectors set up specifically to give early warning about supernovae (imaginatively named the Supernova Early Warning System). How many has it detected since it was first set up a decade ago? None. How many are expected? About 3 per century. So sure, maybe one day we will be able to measure exactly how much light can catch up to neutrinos from a supernova. So far we have exactly one, entirely accidental, point of data.

Let's do the experiments and see how things pan out.

 

[Farsight]

Just exactly how is it going to concertina with a tau of zero? I can imagine a wave function concertinaing at c, but I can also imagine a unicorn, too.

I don't know exactly. Do your own research, look at web pages like this and this.

And please give an example of a helical wave function, particularly as pertains to a particle such as a neutrino.

Again do your own research, search on neutrino and helical. See for example hyperphysics. Or search on photon helical wavefunction.