Saturday, October 15, 2011

Faster than Light?

Last month a team of physicists from the European Center for Nuclear Research (CERN) in Switzerland announced that they had sent a burst of particles across 450 miles to collaborators in Italy -- and they appeared to complete the journey faster than light. According to Einstein's theory of relativity, light establishes the ultimate speed limit for the Universe, and nothing can move faster. My thesis adviser Ed Nather at the University of Texas had a simple rule of thumb for such situations: "Never bet against Einstein".

The experiment itself was conceptually simple. The scientists used the huge accelerator at CERN to produce a bunch of sub-atomic particles called neutrinos, and then beamed them toward a laboratory in Italy where they could be detected. Neutrinos do not interact very much with other matter, so most of them would be expected to complete the journey -- and with something like a large vat of cleaning fluid surrounded by light-sensitive detectors, a small fraction of them could be measured at the finish line. If the distance between the two locations could be determined accurately, the time required to make the trip would reveal the speed of the neutrinos. After repeating the experiment many times, the physicists came to a startling conclusion: the particles took 60 billionths of a second (60 ns) less than it would take light to travel between the laboratories. Light moves about 1 foot (30 cm) in a billionth of a second, so a 60 ns discrepancy corresponds to an error of about 60 feet (18 meters) on the distance between the two locations. Using GPS satellites, they actually knew the distance to within about 8 inches (20 cm), and they also had very precise measurements of the time. So they announced the result and asked for other explanations.

Astronomers were quick to point out a serious problem with the result. If neutrinos really moved slightly faster than light, then those produced in the most recent nearby explosion of a massive star should have been detected about 4 years before we saw the star grow brighter in 1987. In fact, the neutrinos from that "supernova" explosion were detected just a few hours before the light, which is consistent with what we understand about how stars end their lives. So something had to be wrong with the experiment -- and the culprit was probably identified this week by a Dutch mathematician. It turns out that the GPS satellites are moving fast enough in their orbits around the Earth that corrections are needed to the times that they record for the production of the neutrinos in Switzerland, and their subsequent detection in Italy. Ironically, these corrections come from Einstein's theory of relativity and amount to 32 ns on each event, for a total difference of 64 ns -- within the uncertainty of the 60 ns discrepancy observed by the CERN team.

Einstein has stood the test of time. In his one mistake of failing to predict the expansion of the Universe discovered by Vesto Slipher and Edwin Hubble (which he described as the "greatest blunder" of his career), he ended up predicting the "dark energy" that was discovered more than half a century later. So the next time you hear a claim about faster-than-light travel, just remember Ed's rule of thumb.