On hadron colliders, dark matter and black holes

[bartmoss]

Some sort of tunneling effect? Gief stutterwarp drives!

[RPMiller]

RPMiller: The measurements have been aggregated over two or three years. A micro black hole isn't sticking around for that long.

It wouldn't have to. Just long enough to affect the passage of the neutrinos during the test, and if the holes were popping up en mass and then collapsing, it would be even more likely. But as I said, wild speculation and science fiction.

[torstan]

@cantab - the variation of the gravitational field over their travel distance is negligible in this case. The gravity is dominated by the mass of the earth, and that's going to be more or less identical along the path of their journey.

@RPMiller - actually a variation of that is already being suggested. One proposal is that there's an extra dimension that neutrinos of this energy can access that allows them to shortcut the distance. So they aren't travelling faster than light in our 3D universe and relativity remains unbroken. However this is wild speculation. The overwhelming odds are that this is an experimental error with something like the synchronisation.

[RPMiller]

That would make a whole lot more sense, and is far more conducive that to err is indeed human.

[Redrobes]

I don't think that measuring the position of the point above the mountain to an accuracy of 20cm is much of a problem. In the UK you can measure positions with a pro GPS (not our lacky hand held jobs) using differential readings from the standard reference points to about 1mm. You can even do this yourselves with a handheld GPS and download the differential offsets from the Ordnance Survey website in the UK. It would be more accurate than the 5m or so the hand helds have - probably something like 0.5m but I am assured that the pro versions time to a much higher accuracy. Some mapping agency in France/Swiss must have the same set of reference points and differential GPS receivers to do this.

So then you have to know the position of the detectors relative to a reference point on the mountain surface. Well again, given that this is a smaller distance to cover then you can do that pretty accurately now with the modern theodolites. So I would think that its not out of the realms of possibility that you could know these distances to a significant accuracy if you trust the measuring tools. The thing is, most of these measuring tools now use c as a constant in the calculations.

A pair of synced atomic clocks would have an accuracy higher than that needed to get to nano second measurements. I would be pretty sure you could do the radio pulse thing too and account for the positions of the radio emitter and detector relative to the particle detectors. I thought it was possible to sync atomic clocks at a distance if you know their position. Presumably this must go on all the time between all the time standardization agencies. Given that you should be able to see 10 or 12 GPS satellites at once and they all radio in the time stamp and they have to do this with 1mm of light speed accuracy then it cant be hard to sync the two ends and you have the redundancy to check that its correct.

The GOCE satellite was supposed to have mapped the earths gravity profile to an accuracy of about 1km so you ought to know to a pretty reasonable degree of accuracy what the gravity factor is right across the path taken to account for that affecting the clock tick. I am sure they know the moons position and mass well enough to account for it. In any case, does it affect the speed of light that much. Its not like its passing close to some black hole. Its so slight that we cant even perceive the change. Must be many many orders of mag down from something which would add 50nS to the error.

There's plenty of parameters to get wrong tho but the basic ones seem like measurements that can be checked with enormous redundancy.

Doing the same experiment in a completely different test setup and location would be the best thing to try tho. Even if the result does turn out to be a miscalculation, a mistaken measurement or some faulty equipment, it would be real interesting to know what it was.

[bartmoss]

@RPMiller - actually a variation of that is already being suggested. One proposal is that there's an extra dimension that neutrinos of this energy can access that allows them to shortcut the distance. So they aren't travelling faster than light in our 3D universe and relativity remains unbroken. However this is wild speculation. The overwhelming odds are that this is an experimental error with something like the synchronisation.

The synchronization is also the absolutely most obvious source of error and probably the best checked. I absolutely agree that this will turn out to be human or systematic error, and it's not like we never did anything with neutrinos before... But hotdamn I hope this is new physics!

[Jaxilon]

My goodness...micro black holes warping things through time....now I know what's been happening to my brain!!

All this is the beauty of science. That we keep learning and finding out that there is more to learn. The Earth used to be the center of the universe, then the Sun was, now it's something else. Gravity has more than one rule and all that jazz. It's awesome stuff. I kind of hope it's not a human error but another door to greater understanding about how things really work.

[torstan]

I think the odds on this one really are that it's human error. I'm leading a discussion in the department tomorrow about this. I'll post a note of what comes out of it. Looks like we have Nobels wading in on this now - certainly interesting. But yes, this really is science at work. We figure out what we don't know and then devise plans to slowly close out the possibilities. Quite literally:

How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?

In this case the improbable really is very improbable indeed.

[Redrobes]

Theres been some talk that there is a some kind of FPGA based timer which is very non standard and could have some dodgy code in it to get the result. Do you know what that device might be. I saw that CERN has an open hardware forum where they are trying to standardize some of the hardware units being used in CERN generally to make this kind of debugging a lot easier. From what I remember they were looking at getting hardware which did very accurate timing in the list of required devices.

Did you catch the XKCD of this one. Obviously he is betting on human error too. I have no idea as its quite a lot of error to get wrong but human error is the most plausible. But I sure am curious about the result.

http://xkcd.com/955/

[torstan]

Yep, I saw that. It's great