Thursday, February 23, 2012

superluminal neutrinos - not (quite) dead yet.

When historians of science look back on the whole OPERA superluminal neutrino discussion, one way or the other, there are going to be a number of lessons to draw from the experience about how science and science journalism function in the early 21st century.

Yesterday, with "BREAKING NEWS" headlines, Science magazine proclaimed: "Error Undoes Faster-Than-Light Neutrino Results". In that article, "according to sources familiar with the experiment", the whole timing discrepancy for the neutrinos is traced to a bad fiber optic connection to a GPS receiver. The claim from that article is that "After tightening the connection then remeasuring the time it takes data to travel the length of the fiber, researchers found that the data arrive 60 nanoseconds earlier than assumed." Since that's the critical amount by which the neutrinos allegedly arrived too soon for special relativity, that would seem to be the end of the story. Embarrassing for OPERA, but case closed, right? T

Wrong. First, on its face, this seems weird - no one is quoted by name, and the idea that a loose fiber coupling could contribute 60 ns in timing is pretty odd (since that would correspond to something like 18 m of free space optical path). At minimum, the description above must be garbled.

Moreover, the actual email from the CERN director does not say this at all. Rather, it says that OPERA has identified two outstanding issues, one involving an oscillator that provides timestamps for the GPS synch, and an optical fiber connector that brings the GPS signal to the OPERA master clock. The former issue could make the neutrino timing problem worse, in fact. Moreover, the message says explicitly that they are going to take new measurements in May to check these issues. That seems to flatly contradict the news article claiming that they've already done tests. For a detailed discussion, see Matt Strassler's excellent blog here.

Bottom line: as I've said before, the superluminal neutrino result is almost certainly wrong, but the jury is still out on how and why, despite what the Science news blurb says. Believe me, if they knew for sure how this stood, they'd end it with a definitive statement, not stretch this out 'til May.

UPDATE:  Prof. Strassler has the best write-up of this, based on detailed reporting from the European press.   Because of two different, subtle technical flaws, the uncertainty in the OPERA results is bigger than the 60 ns timing discrepancy, meaning (1) the result is not in contradiction w/ special relativity (big surprise), and (2) they need to run with fresh data and the problems rectified to make any more definitive statement.


Massimo said...

Well, sure, sloppy and inaccurate press is what you get when you go out and give press conferences...

Douglas Natelson said...

Massimo, the worry is that sloppy and inaccurate press is often the rule rather than the exception, particularly in these days of declining resources for science journalism.... I'm disappointed that Science would have an all-caps headline for an anonymously sourced, likely inaccurate blurb. They should have higher standards.

Anonymous said...

Wouldn't it be ironic, if after correcting for these two competing effects, that the measurement becomes even more superluminal?

David Brown said...

"... the superluminal neutrino result is almost certainly wrong .." The Rañada-Milgrom effect predicts a GPS timing problem that seems to fully explain the OPERA anomaly (i.e. GRT is slightly wrong).
The Pioneer anomaly is a consequence of dark matter as explained by the Rañada-Milgrom real-or-apparent effect. The real effect or the apparent effect would both support a heuristically distorted model of modified general relativity theory (HDMOGRT), in which the -1/2 in Einstein’s field equations is replaced by -1/2 + Rañada-constant/2, where Rañada-constant is roughly equal to sqrt(60) * 10**-5. The particular value of Rañada-constant comes from the Pioneer anomaly data. From the heuristic model of general relativity there is an anomalous Newtonian approximation for gravitational force (but not non-gravitational force):
Gravitational force = (1 + Rañada-constant) * mass * acceleration, where the dark matter force might be apparent rather than real and is merely a result of the extreme difficulty in detecting dark matter particles — or the altered gravitational force might be real and due to an unknown, weird, or funny distortion in Einstein’s field equations. In either case, the Rañada effect might lead to the valid physical interpretation of M-theory. review of Pioneer anomaly
Suppose that dark matter particles are the explanation for dark matter. Suppose F is gravitational force and the magnitude of the gravitational acceleration a is large relative to (µ * a(0) )/m. Let a(0) be Milgrom’s acceleration constant. We have
F = m * a * ((m * a)/(µ * a(0)))) if and only if
F * ( 1 / sqrt(1 – (2(µ * a(0))/(m * a))**2)) = m * a if only if
Einsteinian-redshift*(1 + dark-matter-compensation-factor/2) = m * a,
provided that 2(µ * a(0))/(m * a) = dark-matter-compensation-factor and we choose physical units in which gravitational redshift = Einsteinian gravitational acceleration due to gravitational force, Therefore, Milgrom’s acceleration law indicates that heuristically distorted model of general relativity theory (HDMOGRT) with either dark-matter-compensation-factor or its equivalent Milgrom-factor is observationally correct. In other words, the Rañada-Milgrom real-or-apparent effect explains Milgrom’s Law. The OPERA team ignored the GPS timing slowdown due to dark matter (or its Milgrom=equivalent). ... BELIEVE IT OR NOT ...

CarlBrannen said...

Try this NY Times article for the best description of the possible problem:

Josh said...

Its certainly hard to believe any result that has no controls for false reading. On the other hand, is there anything physical really preventing faster than light travel? Isn't it possible that the Lorentz transformations that lie at the heart of SR could need re-drafting?