Naked Science Forum
Non Life Sciences => Physics, Astronomy & Cosmology => Topic started by: JP on 22/09/2011 22:31:40
-
Ok, this isn't really a question. It's more of a discussion topic. I don't know if you've seen the news, but apparently scientists at CERN claim to have seen particles (neutrinos) moving faster than the speed of light. This would obviously be a big deal, since it would cause major problems for special relativity.
Here's a link to the news:
http://www.reuters.com/article/2011/09/22/science-light-idUSL5E7KM4CW20110922
And here's a link to tomorrow's talk (streamed live on the web):
http://indico.cern.ch/conferenceDisplay.py?confId=155620
So... are we going to have to rethink relativity or is this a case of an error in the measurements?
-
They don't seem to be much faster. Maybe it's because they have negative mass [;D]
-
I don't think so Geezer. It is probably a question of interactions. Neutrinos are less interacting with the environment than photons. But are they going faster than C? That's the question... [:o]
-
http://io9.com/5843112/faster-than-light-neutrinos-not-so-fast
-
"Light would have covered the distance in around 2.4 thousandths of a second, but the neutrinos took 60 nanoseconds -- or 60 billionths of a second -- less than light beams would have taken."
It reads like it wasn't a direct comparison between light and neutrinos but was based on the theoretical speed of light. Maybe, the affect is due to a local gravitational anomaly (time dilation). There's a lot of rock in that area! [;)]
-
They don't seem to be much faster. Maybe it's because they have negative mass [;D]
If they had negative mass then they would be going backward in time.
Thinks [:0]
If they were going backwards in time then they would have arrived before they left!
Presumably that would have been noticed.
-
Most probably its a slight directional error or geometrical error. I'm not sure of the geometry of the situation the size of the source and detectors but the time error represents a distance of about 60 feet in a total distance of 500 miles.
There is one other interesting factor. Neutrinos change between types all the time this is probably a quantum mechanical tunnelling process. Now there are strong reasons to believe that tunnelling takes zero time and if the particles are in slightly different positions along their track before and after each tunnelling happens it might just be possible to get from A to B a little bit faster.
It may also be an energy thing because there are also some slight suspicions that very high energy gamma rays travel at a slightly different speed to lower energy ones.
-
I hope it is the tunnelling (or other unexpected thing) rather than a mismeasurement - but I am pretty certain it will be an experimental/datahandling error.
-
Maybe the neutrino's had the wind in their back [:P]
On a more serious note: When was the last time that they measured that distance to San Grasso, Italy? It looks to me to be a geographical/geometrical error as Soul Surfer already mentioned..
-
http://imgs.xkcd.com/comics/neutrinos.png
[ Invalid Attachment ]
kudos to randall for being so quick and h/t to swansont
-
I hope it is the tunnelling (or other unexpected thing) rather than a mismeasurement - but I am pretty certain it will be an experimental/datahandling error.
Yeah, it probably is just an error in measurement, but it would be great if it turned out to be real, opening up a whole lot of new physics.
On second though, we can't have anyone overturning the mainstream orthodoxy. Deploy the thought police!
-
There are men in grey suits approaching my door and black helicopters circling above - what have you done to me JP?
-
webcast at cern in 7 mins
http://webcast.web.cern.ch/webcast/
-
Bet they used Google Earth to measure the distance [;D]
(A small error in their measurement of the Earth's curvature would account for the difference.)
-
I had the talk on in the background while working.
From what I could understand, they used very precise techniques borrowed from geodesy to characterize the expected time of flight. Things had to be so precise that they couldn't just use GPS to determine distances. They also accounted for possible GR effects (time dilation), so any error from GR was expected to be several orders of magnitude below the discrepancy they saw.
When they publish their results, it will no doubt be pored over by experts in geodesy to find possible sources of error. At the same time, others will be looking for independent ways to check their results.
And of course, the big problem is that if neutrinos do travel faster than light, we should see neutrinos from distant supernovae arriving years ahead of the light. We don't see this--so that has to be explained.
-
They only have to be a few inches off in their distance calculation to account for the observation. The problem is that there is no direct method to measure the distance and there are too many variables that can affect the result when it is measured indirectly.
If they did it "line of sight" between the tops of two mountains and got the same result, I'd get excited.
I think the more likely conclusion is that we will discover people have been selling real estate that isn't real [:D]
-
Yes, but that's trivializing the effort that went into computing the significance of their measurement. If you believe the experts in geodesy, this kind of measurement is well understood, and it's not a case that there are "too many variables." These measurements are usually painstakingly accurate with (hopefully all) possible sources of error accounted for. I'm clueless on this myself, but I trust they know that they're doing.
I still do think it's an error, but it's likely something subtle. When they fix that subtle point, the measurement will probably be just as precise, but fall into line with what they expect.
On a related point, my knee jerk reaction to these kinds of ultra-precise measurements is always that they can't be possible, since you're talking about precision that boggle the mind. I recall seeing some talks by folks at NIST about ultra-precise measurements of time, down to the attosecond (10-18 seconds) scale! At some point, unless you're an expert in the field, you just have to trust that they know how to account for errors.
-
It's a geometry problem. They are trying to determine the length of a chord from the length of an arc where the chord is very close to the arc. A very small error in the determination of the radius of the arc will result in quite a large error (relatively) in the chord length. If the curvature (or gravitational field) of the Earth between the two points is not quite what they think, it could easily produce quite a large difference in the chord length.
I'm sure their measurements of the transit time were highly accurate, but they are relying on other people's data to determine the distance.
If they could determine the distance by shooting a light beam between two points, they would be in much better shape.
-
Another way of thinking about it is that they have produced a super accurate method of determining the straight line distance between any two points on the Earth's surface. Nobody was able to do that directly before now. Those distances have always been estimated from indirect measurements.
I'm willing to bet we will discover the distance estimates have sufficient uncertainty to account for the observed results. Bear in mind that the geodesic experts are hardly likely to admit that the guys at Cern just proved they screwed up [:D]
Just to up the ante a bit, there is a well known expression that seems to describe this situation rather well,
"Garbage in, garbage out."
-
Most probably its a slight directional error or geometrical error. I'm not sure of the geometry of the situation the size of the source and detectors but the time error represents a distance of about 60 feet in a total distance of 500 miles.
There is one other interesting factor. Neutrinos change between types all the time this is probably a quantum mechanical tunnelling process. Now there are strong reasons to believe that tunnelling takes zero time and if the particles are in slightly different positions along their track before and after each tunnelling happens it might just be possible to get from A to B a little bit faster.
It may also be an energy thing because there are also some slight suspicions that very high energy gamma rays travel at a slightly different speed to lower energy ones.
I would like to think in this day and age that scientists need to be a little smarter before contradicting Albert Einstein. Even if it's a little problem with school geometry.
-
mAYBE ITS A NEWTRINO?
-
On a more serious note: When was the last time that they measured that distance to San Grasso, Italy?
Gran Sasso.
"Gran" --> "Grande" = Big
"Sasso" = Rock.
-
Geezer, it's rather pointless to argue over whether they made a mistake in measuring distance or not, since we don't have their data. (And I highly doubt if we're experts enough to offer any useful comments if we did have their distance measurement data). They claim that by using GPS units placed in tunnels and optical triangulation, they have an accuracy of 20 cm. They also say that's the largest source of error in the measurement, so no doubt it's been checked and rechecked many times. (I'd put even money on the error being somewhere else that they didn't check as thoroughly.)
You can check out that section of their talk, where they (roughly) explain their techniques. Click the link in the first post to check out the talk, and it's on slides 32-33. Actually, even if you don't want to listen to the talk, it's worth checking out just for the cool web interface they have. You can click on each slide and it brings up a picture of the slide along with a video of the presenter during the portion of the talk for which that slide was shown.
But there's always a slim chance they discovered something new, and since they did a very thorough job of checking their work, it will no doubt be taken seriously and rechecked by others who are a bit more expert than us. :)
-
Official article:
http://static.arxiv.org/pdf/1109.4897.pdf
-
Geezer, it's rather pointless to argue over whether they made a mistake in measuring distance or not, since we don't have their data. (And I highly doubt if we're experts enough to offer any useful comments if we did have their distance measurement data). They claim that by using GPS units placed in tunnels and optical triangulation, they have an accuracy of 20 cm. They also say that's the largest source of error in the measurement, so no doubt it's been checked and rechecked many times. (I'd put even money on the error being somewhere else that they didn't check as thoroughly.)
You can check out that section of their talk, where they (roughly) explain their techniques. Click the link in the first post to check out the talk, and it's on slides 32-33. Actually, even if you don't want to listen to the talk, it's worth checking out just for the cool web interface they have. You can click on each slide and it brings up a picture of the slide along with a video of the presenter during the portion of the talk for which that slide was shown.
But there's always a slim chance they discovered something new, and since they did a very thorough job of checking their work, it will no doubt be taken seriously and rechecked by others who are a bit more expert than us. :)
JP - It's not so much a question of distance. It's a question of radius (or angles if you prefer).
The arc length is very short. A very small change in radius will produce a significant change in chord length. If the Earth's surface between the two points turns out to be slightly more curved than the value they used, the real chord length will be shortened significantly. I suppose I'll have to listen to the thing to find out how they determined the radius. If it's based on GPS data, I think they are up a tree.
As you tell us often, extraordinary claims require extrordinary evidence. Mr Ocam tells us that the simplest explanation probably applies. It's very difficult to measure the chord length accurately (unless, of course, you use neutrinos), therefore............
-
Official article:
http://static.arxiv.org/pdf/1109.4897.pdf
Thanks!
-
I went over the paper. There is hardly anything about how they determined the accuracy of the distance measurement. It's nearly all about timing accuracy.
I did find this statement on page 9:
"and by transposing their positions with a terrestrial traverse down to the OPERA detector."
I think that means people with theodelites.
-
Yeah, they said something similar in the talk. I have no idea how much error a "terrestrial traverse" causes.
-
Yeah, they said something similar in the talk. I have no idea how much error a "terrestrial traverse" causes.
Hopefully not much! It would be a bit embarrassing if the whole business was mucked up by a bit of sloppy surveying. Considering how critical the path length measurement is, I was a bit surprised that they don't go into it in a lot more detail. That's likely to be the thing skeptics (like me) pounce on first.
We'll have to wait and see if similar results can be produced elsewhere.
-
They must have a reasonable confidence that their measurement accuracies are sufficient to allow this time difference to be quoted and with modern kit it should be well within a foot over this distance. Let us have a think about other reasons why there might be a small difference.
Firstly from supernova 1987a where a neutrino pulse was measured we have good evidence that neutrinos travel through EMPTY SPACE at very close to the velocity of light. The velocity error observed in the experiment is 4 parts in 10,000. This supernova was around 160,000 light years away so if the velocity error was that great through empty space the neutrino pulse would have come ABOUT FOUR YEARS before the light pulse appeared! Instead it was in effect just a few days before. Just about the time that it would take for the explosion to blow up the star big enough to be recognised as a supernova.
Could it then be that neutrinos travel faster through dense matter than empty space. that is the refractive index for neutrinos is negative. How could this possibly be.
Now solid matter consists mostly of empty space together with nucleons which the neutrinos also have to pass through now the nucleons represent only a tiny part of the matter. funnily enough this is around one part in 10.000 of the size of the atoms so if in effect the neutrinos travel through this material "faster" (I use the inverted commas with intent) this might just be possible.
Could this in effect be done without contravening the normal rules about the velocity of light?
Well maybe because particles are always waves as well and if travelling through a nucleon causes a phase change in the wave it might well be possible that the apparent "position" of the neutrino is changed and moved ahead of where it might be expected to be if the nucleon was not there and this could integrate into a significant time on a long journey through solid material.
This means that the basic laws of physics are not affected but it might have some interesting effects when it comes to modelling pulses of energy flow in condensed objects like neutron stars.
-
Someone says the statistical interpretation of the experiment is wrong:
http://johncostella.webs.com/neutrino-blunder.pdf
-
Someone says the statistical interpretation of the experiment is wrong:
http://johncostella.webs.com/neutrino-blunder.pdf
Thanks Lightarrow! Very interesting.
Reducing 16,000 events with a tolerance of 10.5 microseconds to a probable tolerance of 6.9 nanoseconds obviously requires very careful analysis.
-
Someone says the statistical interpretation of the experiment is wrong:
http://johncostella.webs.com/neutrino-blunder.pdf
Thanks Lightarrow! Very interesting.
Reducing 16,000 events with a tolerance of 10.5 microseconds to a probable tolerance of 6.9 nanoseconds obviously requires very careful analysis.
And the paper says virtually nothing about how they calculate statistical error...
-
It is given that the speed of light is a barrier which is infinitely high. Up to now all experiments has proven this. But if that barrier is not infinitely high than accordance the very basis of quantum mechanics particles have a transmission change (T). For a square potential barrier a good approximation T=exp(-2Ba) where B =((Vo-E)2m/h^2)^0.5, Vo is barrier hight, h must be h-bar. Still the barrier is extreem high but not infinitely. Is this possible ?
-
I think that the refutation of statistical accuracy paper found by light arrow is very plausible and they are pushing their theoretical accuracy far too far
-
I think that the refutation of statistical accuracy paper found by light arrow is very plausible and they are pushing their theoretical accuracy far too far
It was plausible, but it had an error. John Costella has a new paper up (same link) that explains his error and why they were right about their uncertainty...
-
I think that the refutation of statistical accuracy paper found by light arrow is very plausible and they are pushing their theoretical accuracy far too far
It was plausible, but it had an error. John Costella has a new paper up (same link) that explains his error and why they were right about their uncertainty...
The plot thickens [;D]
-
Maybe this also blows away the theory that an object traveling at the speed of light increases in mass?
"as an object approaches the speed of light, the mass increases without limit"
and even though the mass of a Nuetrino is extremely small - it would still increase.. to what degree???
wonder if they can detect any change in mass ( another experiment perhaps!)
But this question of light speed..... is it truly a Limit!? or just that we'v found nothing that exceeds it! ( until now that is! )
Speed and time dilation - perception of time changing - or factual time change when getting to speed of light or greater.. could this be the explanation?
you go faster than the speed of light, and a "time" shift takes place!?
but then again - does time exist? isn't it simply a measurement we devised to keep track of one day to the next, one moment to the next, a means to reference something that took place in the past, and something that will take place in the future.
"Time" is of no consequence to the physics of the universe, it's little more than as said, a means by which WE measure one moment to the next? [???]
-
I have to agree with what is said in the new analysis but I am still a bit unhappy about pushing the accuracy so far and the fact that only the results right at the edges of the 10 microsecond pulse are critical.
I would feel happier if it were possible to extract some details of the fine structure of the main body of the beam (mentioned in the main paper) from the results and by using crosscorrelation show that the timing of this also demonstrated the same time of flight measurements. This would then show that there was not some other unknown process causing any blurring of the results.
-
Maybe this also blows away the theory that an object traveling at the speed of light increases in mass?
"as an object approaches the speed of light, the mass increases without limit"
and even though the mass of a Nuetrino is extremely small - it would still increase.. to what degree???
wonder if they can detect any change in mass ( another experiment perhaps!)
But this question of light speed..... is it truly a Limit!? or just that we'v found nothing that exceeds it! ( until now that is! )
Speed and time dilation - perception of time changing - or factual time change when getting to speed of light or greater.. could this be the explanation?
you go faster than the speed of light, and a "time" shift takes place!?
but then again - does time exist? isn't it simply a measurement we devised to keep track of one day to the next, one moment to the next, a means to reference something that took place in the past, and something that will take place in the future.
"Time" is of no consequence to the physics of the universe, it's little more than as said, a means by which WE measure one moment to the next? [???]
If this were true why is it that atomic clocks can measure time dilation due to the strength of gravity varying. The latest with an accuracy of 18 decimal places can measure as little a one millimetre in difference in height. If time wasn't 'real' all clocks would measure the same going rate everywhere.
-
If the neutrinos were going faster than light then they would be going backwards in time. This would allow them to arrive before light(?). The time shift would be distance dependant. It would be interesting if a similar experiment were carried out elsewhere but at a different distance. Seems to me the effect is most likely due to some unaccounted for error.
-
I still say they managed to fubar the distance, for no other reason than it appeals to my sense of the absurd [;D]
-
I don't think that anything anomalous does happen on the normal atomic scales but things could very well be different down in dense material (nucleons etc) down at the scale defined by the strong interaction. This may be one of the very few ways that we could measure these effects because it involves scales of 10^-15 metre and times around the time it take light to travel that distance ie 10^-23 second. Remember many cosmologists consider that at very close to the instant of the big bang the velocity of light is in effect infinite.
One of the other things that I have been considering is that quantum mechanical uncertainty is involved in these interactions and this greatly increases the "random noise" in the experiment. However I cannot see how this could introduce any bias, only a great deal more variation within which this result would be a statistically typical result. One of the ways this might be tested is to take say four our five subset groups of the results taken at various times and analyse them independently to see if they all individually show traces of this bias.
-
is it possible that the true value of c is determined by the speed at which neutrinos travel and that photons travel slightly slower than c ?
Does this mean that we have to redefine the meter ?
-
I don't think so for either case. the speed of light pops out of maxwell's equations and is fundamental to SR - so if neutrinos go faster then they do; but it does not affect c. the metre is merely an arbitrary distance - the fact that it is tied in the speed of light in its definition does not mean that it depends on the speed of light for its measurement. as a mad example - if the great flying spaghetti monster (http://www.venganza.org/) changed by fiat the speed of light to 300,000,000 m/s exactly - we would not change the length of our metres, we would merely redefine the metre to the distance light travels in 3*10^-8 sec (or maybe something else entirely, less at the whim of a mad god)
-
There will be found to be something weird involved here, either a mistake of some kind, or 'tunneling' 'backward time travel' 'distance shrinking' whatever. But I'm willing to bet that it won't change 'c' as the constant it is, also that it won't mean that we ever will be able to 'time travel' backwards..
-
Ocam would also tell us that gravitational forces are speeding the acceleration of the expansion of the universe. But since we know what gravity is and that it could cause something very much like we are experiencing if the sources were in the correct positions and they could be the hiding places of the missing mass it would be too easy. So let’s come up with some negative gravity or dark energy and think of something else whimsical to answer the mass question. That’s the way we roll!
Too cynical?
-
Presumably the neutrino beam was made up of both neutrinos and anti-neutrinos?
If so then it is certainly conceivable that the anti-neutrinos were going backwards in time which would make them appear to be going faster than the speed of light.
-
Mike - the paper only mentions neutrinos (but it would as the timing would be identical). the neutrinos were formed by decay of pions (and kaons)
π+ -> μ+ + νμ
π- -> μ- + antiνμ
Frankly I cannot find which form was happening - and as neutrinos are uncharged they could be their own anti-particle - does it really matter.
-
Mike - the paper only mentions neutrinos (but it would as the timing would be identical). the neutrinos were formed by decay of pions (and kaons)
π+ -> μ+ + νμ
π- -> μ- + antiνμ
Frankly I cannot find which form was happening - and as neutrinos are uncharged they could be their own anti-particle - does it really matter.
If they are their own antiparticle then no it does not matter. However, if they are not their own antiparticle, if they have mass for example then conceivably they could be going backwards in time in which case it would matter, a lot.
-
They can be shown going backward in time on a feynmann diagram and quantum states of particle and anti-particle are interchangeable under cpt reversal - but they do not undecay and they do not travel backwards in time in a lumpen reality sort of way.
AS an example - the positron in a PET scan can be perfectly calculated in qed etc by thinking of them as an electron travelling backwards in time - but to make a picture of the inside of your head they are a positron going forward in time