Monday, September 19, 2011

Tachyons At Opera?

Rumor has it (from an anonymous source on September 15 here) that the OPERA neutrino detector at CERN is seeing a six sigma signal (above the usual rule of thumb standard for "discovery" in physics) of neutrinos traveling at more than the speed of light, which means that they would be tachyons, the name for particles that travel at more than the speed of light. The experiment's main purpose was to detect tau neutrinos. Data on the speed at which neutrinos travel is simply a bonus (although neutrinos speed, correlated with missing energy in the source reaction, in theory, ought to provide some direct constraints on neutrino mass).

Tommaso Dorigo who blogged this rumor, and Lubos, who blogged Dorigo's gossip, like all good physicists, are deeply skeptical of the result on the theory that extraordinary claims require extraordinary proof. This is only the second credible experiment (the other looked at neutrinos leaving a 1987 supernovae, which has many more uncertainties involved, for example, concerning the distance and the nature of the underlying source), ever, in the history of the universe, to see anything that might be a tachyon at statistically significant levels. There are good theoretical reasons to doubt the result, despite the fact that some quantum mechanical formulas like the Feynman photon propogator in QED recognize that there is a slight statistical probability of particles traveling at more than the speed of light at short distances incorporated into the equation that are interpreted by some as evidence of a slight amount of non-locality in space-time.

A delay in publication of the results or a press release is no doubt due to the frantic efforts of CERN scientists to find some other explanation of the results (the fluke could be as simple as a slight mismeasurement of the 732km distance from the neutrino source to the point at which they were detected, or a clock miscalibration in measuring events about 3 milliseconds long, or an underestimate of statistical uncertainty) so that they don't look stupid when someone else discovers one. But, the experimental setup is generally sound and much of the theoretical analysis is relative straight forward since it does not involve the arcane and complex QCD background calculations that go into making sense of the debris of high energy proton-proton collisions, so this is not an easy result to hand wave away.

The most plausible answer would be that the neutrino impacts that look superluminal didn't actually come from the neutrino generating source that was inferred, and instead came from nuclear reactions either in the Sun, or radioactive isotypes in the Earth (perhaps related to the geological activity of volcanos in central Italy dredging up an excess of radioactive isotypes from the center of the Earth aligned with the neutrino beam), or from some unaccounted for man made source like a nearby smoke alarm with a radioactive isotype that is decaying in it. A failure to properly account for these kind of neutrino backgrounds has already been suggested in the direct detection of dark matter experiments showing a positive detection that have been conducted so far.

But, information on the direction in which the debris from the neutrino impact travels, the momentum of that debris, and statistical tests ought to be able to rule out those kinds of neutrino backgrounds very effectively in a series of results numerous enough to provide a six sigma signal.

UPDATE: The comments at Resonaances and post at Dorigo's blog regarding the rumored results have both been taken down.


Andrew Oh-Willeke said...

More analysis here, particularly of the 1987 supernova event and noting that the clocks used in this experiment may have been less precise than would be optimal since they weren't designed to measure neutrino speed as their primary purposes.

Andrew Oh-Willeke said...

Kea is thrilled and digs into the idea.