Have Scientists Found Two Different Higgs Bosons?

Yesterday researchers at the Atlas experiment finally updated the two-photon results. What they seem to have found is bizarre—so bizarre, in fact, that physicists assume something must be wrong with it. Instead of one clean peak in the data, they have found two an additional peak.* There seems to be a Higgs boson with a mass of 123.5 GeV (gigaelectron volts, the measuring unit that particle physicists most often use for mass), and another Higgs boson at 126.6 GeV—a statistically significant difference of nearly 3 GeV. Apparently, the Atlas scientists have spent the past month trying to figure out if they could be making a mistake in the data analysis, to little avail. Might there be two Higgs bosons?

But let’s not let this intriguing blip distract us from the original scent of new physics. Back when the preliminary data seemed to show that the Higgs was decaying into two photons more often than it should, I wrote that it could be “a statistical blip that would wash away in the coming flood of data.” But more data has now arrived, and the blip hasn’t gone anywhere. The Higgs boson continues to appear to be decaying into two photons nearly twice as often as it should.

undoubtedly due to a calibration error

The difference of the mass measurements between the two channels is 3.0 ± 0.8 (stat) +0.7−0.6 (sys) GeV, corresponding to about 2.7 standard deviations, using Gaussian pdfs for systematic uncertainties. A more conservative treatment of the systematic uncertainty related to the mass scale, using rectangular pdfs for parts of the error, reduces the difference to the level of 2.3 standard deviations.

V50, I hope you're not knocking the Resonaances blog, which is accurate and well-informed, and IMO one of the top five sources of particle information on the web. He makes clear, I think, that the "two peaks" are the best fits in different channels, 123.5 GeV for the ZZ channel and 126.6 for γγ, undoubtedly due to a calibration error.

@K^2: Other decay channels, their branching fractions (which agree very well with the SM) and the angular distribution to determine its spin (which indicates spin 0).