Original URL: http://www.theregister.co.uk/2011/10/18/reviewing_the_opera_papers/

OPERA review serves up a feast for physics geeks

Superluminal neutrinos and silly science writers

By Richard Chirgwin

Posted in Science, 18th October 2011 21:26 GMT

Let’s get the “big news” out of the way first: there’s a lot of excitement due to one paper published on Arxiv.org, which asks whether the CERN OPERA experiment – the one that seemed to detect superluminal neutrinos – took into account the “satellite reference frame” in its calculations.

So the short version of the paper is this: did the calculations conducted on the OPERA data treat the clocks on the GPS satellites as being stationary with respect to the detector (more precisely for The Register’s crowd-sourced scientific sub-editors, I mean “in the same reference frame as the detector”)? If so, van Elburg argues, it would be a source of error, because the satellites aren’t stationary; and the error he predicts a worst-case correction of 64 nanoseconds.

The media excitement is comprehensible for several reasons: the paper, by Roland van Elburg of the University of Groningen, yields a calculation quite close to the 60 nanoseconds that the neutrinos seemed to travel too fast; it would therefore restore relativity (and give journalists an opportunity for some “you was wrong!” schadenfreude).

It also suits journalists’ sense of narrative: Einstein’s theory used to prove that Einstein was right.

The most sensible discussion of this hypothesis I've seen is on the Bad Astronomy blog, here.

What the sudden leap for “Einstein still right” headlines fails to convey is the ongoing scrutiny being applied to the OPERA experiment. It hasn’t been called into question by one paper; what’s happening is an intense scrutiny that’s mostly passing unnoticed.

By my count, there have been 47 papers published at Arxiv.org discussing the OPERA results one way or another. So why just focus on one? Instead, let’s scan a few of the ideas, discussions and quirks of this “gold standard peer review” (as one reader described it recently).

Causality safe, for now

One of the first things that got everybody excited when the OPERA results first hit the wires was that the possibly-superluminal neutrinos traveled “backwards in time”. By my admittedly-poor understanding of relativity, this was mostly a journalistic construct, since in a human time-frame, the neutrinos still arrived after they left.

Nonetheless, time travel grabs the attention like few other topics, apart perhaps from climate conspiracies and crop circles. So please don’t direct any hate mail to me or to the physicists for telling you it ain’t so, at least not yet.

According to a German-Hungarian group of physicists, even if the neutrinos could travel faster than light, they can’t send information backwards in time. Their paper reminds us all that the neutrino beam would need, somehow, to be modulated. For us to be satisfied that a bit of information has been transferred, the trailing edge of the modulated pulse, not the leading edge, has to be observed fast enough to satisfy the “faster than light” requirement.

It may be, they admit, that a setup could be devised to defeat this, but at the moment, causality is safe from messages-from-the-future.

Interpreting the OPERA signal

If you’re in the mood for statistics, Antonio Palazzo of Technische Universitat Munchen has the paper for you.

It’s worth a look, even if only to dispel the pop-science idea that the neutrino observations were as simple as switching on a beam near the LHC and waiting for it to arrive at Gran Sasso. Rather, the researchers looked at a huge number of interactions between neutrinos and their detectors, applying a statistical analysis to derive a “probability density function” of the neutrino emission times.

That density function is designed to identify the point in two somewhat “smeared” waveforms – the proton events that produced the neutrinos, and the neutrinos detected at Gran Sasso – for which the experimenters can state “these neutrinos were definitely produced by the CERN proton beam at time X”.

Palazzo’s question is whether the right statistical techniques were applied: “it seems that the single waveforms are first summed together and then their sum is normalized to the total number of neutrino interactions in OPERA … such a procedure, if effectively adopted in OPERA, is questionable”, the paper states.

To help resolve this, Palazzo has asked that the OPERA researchers publish the timestamps of the experiment’s 16,000 detected neutrino interactions, along with those of the associated proton waveforms.

Walter Winter, however, doesn’t agree with Palazzo, asserting in this paper that “possible smearing effects … do not change the OPERA results”. That’s not, however, his main concern: Winter also questions the popular (among physicists) assumption that the superluminals must be “sterile” neutrinos.

His re-analysis of the reported data suggests that the “sterile superluminal” theory “can probably be ruled out”: some non-sterile types of superluminal neutrinos are, he suggests, necessary to explain the number of superluminals apparently observed.

The superluminal neutrino laser

A personal favourite of mine in the numerous attempts to explain the OPERA results is this paper by Rafael Torrealba at the University Centro Occidental in Venezuela.

In spite of some translational difficulty, Torrealba proposes the neat idea that the CERN experiment has actually invented a kind of “neutrino laser”: “the starting point of neutrinos could be shifted in time or driven by stimulated emission, as happens for an ultrashort pulse LASER traveling through an amplifier plasma with an initial population inversion”.

“In other words”, he writes, “the traveling distance of the stimulated neutrinos is shorter than that of the not stimulated ones.”

The world still turns

GPS relativity isn’t the only “mundane” source of possible error: two authors have independently asked whether the Earth’s rotation was properly accounted for in the OPERA analysis.

Markus Kuhn of the University of Cambridge (here) asks whether the effect is properly accounted for – while admitting that it would only have a 2ns influence on the reported result, insufficient to invalidate the apparently superluminal trip.

Dominique Monderen – whose paper doesn’t give an affiliation, proving that this review process truly is open to all – also queries the impact of the Earth’s rotation, noting that “the distance [of the neutrino trip] and the time of flight are measured in two different inertial time frames.”

Distance, she says, is measured in a static time frame, but the time of flight experiences Earth’s rotation.

Any of the discussions hitting the wires over at Arxiv.org could be the answer, or none of them might be.

However, before seizing on any single item as being a candidate for refuting the OPERA results, science journalists – who really should know better – would do well to keep a couple of things in mind.

First, the publication of the OPERA results at such an early stage was designed to attract exactly this kind of public scrutiny.

Second, nearly all of the explanations, questions and refutations will themselves need to be reviewed before they can be accepted. ®