Blog – Of Particular Significance

[2:40 Paris time: UPDATED ]

The ATLAS experiment (in a talk by Olivier Arnaez at the Higgs Hunting workshop in Orsay, France that I’m attending) has made public its search for Higgs particles decaying through real and virtual W particles to a lepton (electron or muon), anti-lepton, neutrino and anti-neutrino.  [CMS made its result for this search public on July 4th at the big presentations at which discovery of the Higgs-like particle was announced.]  The importance of this search is now no longer to tell whether the Higgs-like particle exists — we are confident that it does — but to try to measure how often these particles are produced and then decay in this particular way.

This is a very difficult measurement, with a small signal and a complex and large background, and it only allows a measurement of a rate of the process; the mass of the particle producing an observed excess cannot be determined to better than 30-50%, because the neutrino and anti-neutrino are not observed, and the lepton and anti-lepton are not enough to figure out the mass of the parent particle.  Meanwhile, small amounts of data can play tricks.  After last summer’s data, ATLAS saw a sign of something.  By March, with the full 2011 data, they said they didn’t.  Now, including 2012 data, they see it — in fact, they see slightly more than expected for the simplest possible type of Higgs particle (“the Standard Model Higgs) of mass 125 GeV/c².

More details to follow shortly, but the news is:

• ATLAS observes in 2012 data a 3.1 sigma deviation from the no-Higgs hypothesis
• Combined with 2011 data (which showed almost no excess) there is now a 2.8 sigma deviation from the no-Higgs hypothesis
• the excess looks like a Higgs particle; it is quite consistent in its transverse-mass distribution with the hypothesis of a Higgs-like particle at a mass of 125 GeV/c²
• the fit to the signal is 1.4+-0.5 times larger than the Standard Model prediction for a Higgs particle of mass of 125 GeV/c² (though one should not forget that this measurement isn’t good for telling us the mass; just by itself, it would also be consistent with a lighter Higgs particle produced with an expected rate or a heavier Higgs particle with a much lower than expected rate.)
• this in turn is in good agreement with what ATLAS observes in the two-photon and four-lepton searches, both of which are a bit higher than the prediction for the simplest Higgs with mass of 125 GeV/c²
• this also is consistent specifically with the prediction, which holds for most expected types of Higgs particles, that the ratio of the strengths of the interaction of the new particle with W and Z particles is equal to the ratio of the masses of the W and Z.

These results are also roughly consistent, within the uncertainties, with those from CMS, which sees a slightly smaller production rate than expected, something like 0.6 +- 0.4 (don’t quote me, I’m reading it off a plot) times what is predicted for a simplest Higgs of mass 125 GeV/c².

NOTE ADDED: Did I mention this is a difficult measurement? The excess in the 2012 data looks a lot like a Higgs particle signal, but a mis-estimated background would look pretty similar (though admittedly it would have to be a big mis-estimate — but then again, the signal in 2012 is itself more than twice as big as expected), and so we’re reliant on the estimates of the systematic uncertainties on the backgrounds given by the experimenters.  There are differences between the 2011 and 2012 analysis techniques (improvements, surely) that I don’t understand yet.  So… an impressive result, but my own view of it is still a little murky.  I’m glad this excess wasn’t essential for the claim of Higgs particle discovery.