Comment by nickolai | Hacker News Reader

Comment by nickolai | Hacker News Reader

Similar results seem to have already been obtained in the past

http://en.wikipedia.org/wiki/Neutrino#Speed

QUOTE

In the early 1980s, first measurements of neutrino speed were done using pulsed pion beams (produced by pulsed proton beams hitting a target). The pions decayed producing neutrinos, and the neutrino interactions observed within a time window in a detector at a distance were consistent with the speed of light. This measurement has been repeated using the MINOS detectors, which found the speed of 3 GeV neutrinos to be 1.000051(29) c. While the central value is higher than the speed of light, the uncertainty is great enough that it is very likely that the true velocity is not greater than the speed of light. This measurement set an upper bound on the mass of the muon neutrino of 50 MeV at 99% confidence.

/QUOTE

The value looks awfully like what we have in front of us today, but the uncertainty was too big to investigate further

EDIT : actual paper :http://arxiv.org/abs/0706.0437

Experimental error seems likely to me once again. 10ns at c is 3 meters, over 730km + uncertainty in time of particle generation and decay + detector uncertainty...I don't buy it. From the project site: http://operaweb.lngs.infn.it/spip.php?rubrique41

OPERA needs an intense and energetic beam of muon neutrinos traveling a distance of hundreds of kilometers to seek for the appearance of oscillated tau neutrinos. A beam of this type is generated from collisions of accelerated protons with a graphite target after focussing the particles produced (pions and kaons in particular) in the desired direction. The products of their decays, muons and neutrinos, continue to travel in generally the same direction as the parent particle. Muon neutrinos produced in this way at CERN cross the earth crust reaching OPERA after a 732 km journey.

>I don't buy it

I didn't mean to try to 'sell' anything, though I agree that my comment was needlessly biased here. Revised it accordingly.

On a side note, it's pretty wild seeing history being written in Wikipedia. The BBC site has just been referenced on that part of the Neutrino article.

> ... found the speed of 3 GeV neutrinos to be 1.000051(29) c.

What does the "(29)" in "1.000051(29) c" mean?

Repeating. It's 1.000051292929292929...

edit: sorry if that's incorrect in this case. "1.000051(29)" is one of the formats of indicating repeating decimals, per http://en.wikipedia.org/wiki/Repeating_decimal#Notation

I'm not sure, but I think it represents uncertainty over the last digits.

So 1.000051(29) would be equivalent to 1.000051 +- 0.000029.

Yes, that is what it means. :)

That can't be it because the error wouldn't straddle the speed of light. It would put the range at 1.000022c to 1.000080c, both of which exceed c, so they wouldn't have written it off as error.

But that doesn't indicate a hard error bound; this is a statistical matter, after all. Most likely it's just an expected standard deviation.

It indeed is the standard deviation.

A single standard deviation is not enough, normal procedure is to allow up to three standard deviations (or more, sometimes five), which puts the measurement in a range consistent with models.

Those figures are being provided because they were measured, but they are within the expected experimental error for the device.

http://www.numericana.com/answer/rounding.htm#significant

That means that the last two digits of the number 1.00005129 are uncertain.

For example, they measured 1.00005129c, but maybe their uncertainty is +-0.0000001c. The last digits are unreliable, but are the measurements they got

EDIT: below, its claimed that those digits ARE the uncertainty. That quite possibly is it; my answer is from what i remember from uncertainty in my 1st year Physics course.

This article (http://www.google.com/hostednews/ap/article/ALeqM5in1T5nvGNc...) mentions previous results and that they were thought to be due to measurement error, but claims that these readings had a speed of c + 60 nS, with a measurement margin of error of +-10 nS.

Reply on news.ycombinator.com