Fermilab is setting up to do it, too. They should be up and running within a year (though this result took three years of data; it would probably take a similar time for another experiment).
Actually, Fermilab did a similar experiment before, and got about the same result that CERN did, but the errors on that earlier experiment were higher, such that the result was not really statistically significant.
It may be possible to improve some of the delay measurements that fed into the older MINOS result, allowing the use of all the already-collected data for a preliminary result with an intermediate precision on the 9-12 month time scale. This work is underway, and much more will be known in the coming weeks and months. Also underway are upgrades to the beam and detector timing systems, which will unfortunately take several months to prepare, which happens to be the amount of time remaining before a one-year accelerator shutdown at Fermilab. (The one-year shutdown will allow for some essential upgrades for on-going and future experiments at the lab). Nonetheless, the goal is to have as many of these new timing systems in place before the accelerator downtime begins so that relative offsets can be measured and applied to older data. This would allow for a better, but still not full precision, result. Finally, a useful data sample can be collected in the first year after the shutdown period ends, and this data would be collected with all the timing-related bells and whistles in place. The resulting precision would be comparable to or better than the OPERA precision, and the measurement itself will have many more redundancies built-in to add to its convincing-ness, regardless of the answer. This full-precision result would be ready no earlier than the end of 2013, according to current accelerator plans.
The T2K experiment in Japan can also make this measurement. They have the disadvantage of having a shorter distance over which the neutrinos travel, by a factor of 2.5. Thus, their ultimate precision on the velocity will be somewhat worse, but there is no reason why it can’t be plenty good enough to test the OPERA claim, which is large. However, it may take a similar amount of time for them to accumulate sufficient numbers of neutrino interactions at the far detector.
Assuming, of course, that they have the same amount of error in their baseline distance and clock synchronization.
EDIT: Oh, and the CERN folks themselves are probably looking into more ways to double-check their own distance and clock synchronization.
True. And, this is a reasonable assumption, at least at the 2.5x level that is relevant. The Cs and/or Rb clocks and the GPS systems face the same technological limitations in both experiments. The baseline distance error should be limited to a few tens of centimeters in both experiments. What remains are the uncertainties in the subsystem timing offsets, which will differ between the two experiments and for which I would guess T2K may have the (slightly) larger errors in the end.
I don’t know that it is a reasonable assumption. The European experiment had 20 cm of position error at the detector end, but only 2 cm at the source end. If the Japanese (or anyone else) can pin down both ends to the precision that CERN got at the source, their distance error would be down by an order of magnitude.
Given the extraordinary claim of CERN, how many others need to confirm the results before it is accepted as fact?
Enough experiments to give predictable results, and preferably using at least one set of entirely different experimental parameters. (different source and target technologies, and at a minimum two verifications of the distances involved.)
I think this one is worth an unmanned space program emitter at least as far as Mars. Not sure how to set up the timing on that one, but it can be done. Building a detector out that far would take a moon base.
Tris
more
I understand this article to say that the original researchers are aware of the criticism and claim they adjusted for this effect, but that they did not explain it as well as they could have in the original paper.
Is that right?
Of course, confirmation of a result like this isn’t a discrete matter: It’s not like it’s unconfirmed, and then there’s one more experiment, and click, it’s confirmed. Nor is it all about the quantity of experiments: Quality counts, too.
That said, there are a few things I’d like to see. First, instead of doing the timing at the source and the detector, do the timing at two different detectors, one near and one far: This will cancel out any systematic errors in the detector readout (the MINOS experiment at Fermilab already did this).
Second, put sources on both ends, so you can perform the experiment both from A to B and from B to A. This will cancel out any errors due to clock synchronization.
Third, if this effect is real, then some explanation is necessary for why it didn’t show up in the SN1987a neutrinos. Presumably, there’s an energy dependence, since these neutrinos were orders of magnitude more energetic than the supernova ones. I’d like to see a decent theoretical model for this effect that’s consistent with the two results.
Finally, once you have such a theoretical framework and calibrate it with the supernova results and results like this, it should be able to predict the speed for neutrinos of intermediate energy. Run experiments at those energies to test it. Plus, of course, other experiments to test whatever other predictions those models make.
Actually, it might be easier to put the detector out in space and the emitter here: Neutrino detectors need to be big, but they’re not all that complicated. You might be able to install an Ice Cube-like detector in one of Mars’ polar ice caps more easily than building an accelerator out there.
And a space-based experiment would also have the advantage that you could run a light beam along the same path, to compare the neutrino speed directly to the speed of light.
But this distance error would then be (even more) negligible in the final result. That’s why I said “limited to a few tens of centimeters in both experiments”. Any error of this size or smaller will be sub-nanosecond, and all other errors combined will be significantly greater than that (in terms of the quadrature sum). So, an order of magnitude better distance measurement doesn’t translate into an order of magnitude better velocity measurement. The distance measurement just needs to be “good enough”, and both will be.
What’s the shortest distance over which lightspeed has ever been measured? Could it be that the SOL can vary over very short distances/ very brief times, but always converges on the standard value over any macroscopic scale?
Well back when i was teaching high school physics, I turned on a flashlight and my students timed how long it took them to see it. They all agreed on one value (0 seconds), but I admit our error bars were pretty big.
While the two MINOS detectors are generally of the same design, the electronics and timing systems are very different, so for the existing MINOS time-of-flight measurement, essentially none of the timing errors canceled (and the detector timing errors were in fact the driving errors on that measurement). Changing the experimental set up so that the errors do cancel out is, as you say, exactly the right thing to do.
Yes. But that Nature News article shows that this particular criticism is evidently still evolving. I would welcome further description of the methodology from the OPERA folks, but I don’t know that it will ever be detailed enough, which is why independent tests will be key. As luck would have it, the NuMI neutrino beam from Fermilab to MINOS starts and ends at latitudes that are nearly exactly opposite to the starting and ending latitudes of the CNGS-to-OPERA line, so if there is an effect here from gravity, it will have the opposite sign in MINOS.
Cool news! Doubtful of a gravitational influence, THAT would impact the models that are well proven FAR worse.
That said, the models that predict such events are rather sparse AND vague and needing refinement. THIS could provide them with such information, if validated by other experiments.
For, without duplication, it is nothing.
As for the questioner, no, the paper that they released showed all common and moderately extraordinary refutations and rather hinted at requests to disprove the measurement.
My overall felling gained from their original paper was one of bewilderment and consternation. Hence, the WAY it was reported in the paper.
Rather a bit of, “I don’t trust THESE results and want SOMEONE to shoot them full of holes, as our BASIS of physics is a fair amount out the window”…
[QUOTE=secret message]
THAT FAR AND THIS WAY THESE SOMEONE BASIS
[/QUOTE]
I don’t get it.
I always thought I was the only one who did this! Do you look for acrostics in newspaper columns, too?
I was poking around and came across John Moffat’s paper (http://arxiv.org/PS_cache/arxiv/pdf/1110/1110.1330v3.pdf ) on the OPERA results which h just went up yesterday. Now this is the guy that’s all about modified gravity so of course his paper focuses on a modification of SR.
So that’s neat. Basically since SN1987a was far enough away that the term influencing matter particle velocity drops to zero, but the smaller distances of the OPERA experiment allow it to appear.