Is particle physics, like beauty, in the eye of the beholder? You would be forgiven for thinking that now that two teams have analysed data from Fermilab's Tevatron collider and come to the exact opposite conclusion about whether that data hints at a new particle.
A task force is being formed to figure out the discrepancy, but the final arbiter may be the Large Hadron Collider in Switzerland, which will ultimately collect more data than the Tevatron.
In April, members of the Tevatron's CDF experiment reported finding a curious signal in the debris from eight years' worth of collisions between protons and antiprotons. The signal hinted at the existence of a particle that was not predicted by the standard model, the leading theory of particle physics. Theorists scrambled to come up with possible explanations, writing dozens of papers on the topic in the following weeks.
Last week, evidence for the signal, or "bump" in the data, seemed to get even stronger. The CDF team reported that it had analysed twice as much data as in April and had still found the bump.
But now, a rival team performing an independent analysis of Tevatron data has turned up no sign of the bump. It is using the same amount of data as CDF reported in April, but this data was collected at a different detector at the collider called DZero.
"Nope, nothing here – sorry," says Dmitri Denisov, a spokesman for DZero.
Different detectors
When the CDF collaboration came out with its result in April, DZero researchers spent a couple of days doing a quick check of their data and saw no bump. But to make sure they were comparing like with like, they spent the next two months painstakingly checking that their analysis resembled that of CDF's as closely as possible.
Today, they are reporting that their analysis shows no bump. "We're basically excluding a signal at well over the 95 per cent confidence level," Denisov told New Scientist. The result shows "good agreement with the standard model".
"Now of course the most interesting question is where are these differences coming from?" he says.
The fact that the detector is different should not come into play, he argues. "It would be really puzzling if it's a physical process that's supposed to exist in nature, and one experiment sees it and another doesn't," he says. "Protons and antiprotons don't know what detector they're colliding in."
Modelling issue?
He does not know what is causing the discrepancy but suspects it may be due to differences in the models that each team uses to describe the data. The studies look at how often collisions between protons and antiprotons produce a W boson, which transmits the weak nuclear force, and a pair of jets of subatomic particles called quarks.
The CDF team found an unexpected abundance of these events – a W boson and a pair of jets – where the mass of the jet pair was about 145 gigaelectronvolts, suggesting that a new particle of that same mass was created (see graph). But other events involving different combinations of particles can mimic the signal of a W boson and a pair of jets. If the CDF team incorrectly modelled the number of those "background" signals, it might make it look as if there were a bump in the data, says Denisov.
Rob Roser, a CDF spokesman, acknowledges that "it could be a modelling issue", but says it is too soon to discount CDF's result. "It's disappointing that the peak didn't just jump out at them too," he told New Scientist. "But just because it didn't doesn't mean there's not something there."
Task force
Roser has not yet had time to look carefully at DZero's paper (pdf), which has been submitted to Physical Review Letters. But he says a cursory look suggests that the discrepancies between the two results may not be as large as it seems. The CDF team estimated that the potential new particle is produced at a certain rate in proton-antiproton collisions, but there is some uncertainty in that rate. If the real rate is at the lower end of CDF's range of possibilities, for example, the discrepancy between the two teams' results is smaller.
"I think we have more work ahead of us before we understand what is going on," says Roser.
A task force, made up of members of both experiments, along with Fermilab theorists Estia Eichten and Keith Ellis, will now try to get to the bottom of the discrepancy, a process that could take months, Roser says.
Roser says the CDF team has only analysed 70 per cent of the data it expects to collect by September and will continue crunching the numbers to see if the bump appears in the full data set. But Denisov says DZero is satisfied with the analysis it has already done. "We do not plan to continue the analysis with the same rigour as we did over the last two months," he says. "We are concentrating on many other things."
Even if the task force's investigation proves inconclusive, both Denisov and Roser say the Large Hadron Collider at CERN will eventually collect enough data to settle the question of whether there is a bump at 145 GeV.
They disagree over when this will happen, though. The LHC collides particles together at higher energies than the Tevatron, which produces more debris in which a new particle could be lurking. This leads Denisov to estimate that the LHC will collect enough data to look for the first evidence of the bump in the next few months.
Roser points out that the LHC collides protons together rather than protons and antiprotons, as happens at the Tevatron. That means that even though it produces some collisions between quarks and antiquarks – which could be important in reproducing the bump – it generates them in smaller numbers than the Tevatron. So Roser estimates it could take the LHC more than a year to make its ruling.
So are the two teams going to be arguing bitterly until then? "It's been a little bit tense over the last few days," laughs Denisov. "But I think it's friendly. I would say it's like two sports teams competing."
"I think it's going to be fun," says Roser. "This is science in progress."
Source New Scientist
No comments:
Post a Comment