Things that go bump in the light

A packed house in the CERN auditorium awaits a presentation from Jim Olsen from Princeton University on December 15, 2015. That presentation summarized the 2015 results from CMS, including an unexpected bump (see below). Photo: CERN

A packed house in the CERN auditorium awaits a presentation from Jim Olsen from Princeton University on December 15, 2015. That presentation summarized the 2015 results from CMS, including an unexpected bump (see below). Photo: CERN

In the latter half of 2015, after a nearly three-year hiatus, the LHC delivered collisions at an unprecedented energy of 13 trillion electronvolts (TeV). On December 15, as a scheduled winter maintenance shutdown began, scientists from both ATLAS and CMS presented the first results using that cumulative 2015 data set. Among those results were modest excesses in observed photon pairs, which, perhaps most unexpectedly, popped up at the same energy in both experiments. Such a coincident excess could be caused by a previously unknown particle decaying into photon pairs or could simply be a coincidental statistical fluke.

This unexpected bump in the masses of photon pairs was seen by both the ATLAS and CMS experiments. According to the Standard Model, nothing heavier than a Higgs boson decays solely into two photons, so might this bump represent a new particle?

This unexpected bump in the masses of photon pairs was seen by both the ATLAS and CMS experiments. According to the Standard Model, nothing heavier than a Higgs boson decays solely into two photons, so might this bump represent a new particle?

Broadly speaking, searches for new physics at the LHC can be lumped into two approaches. In one, physicists select collision events that have a ready prediction in the familiar Standard Model and look for anything out of the ordinary. In the other approach, physicists search for a telltale signature corresponding to a specific model of new physics. The former includes the search for new physics in photon pair events. In this case, physicists identify events that have two energetic photons in them and calculate a combined “mass” for them. If the two photons resulted from the decay of a heavy particle, such as a Higgs boson, the calculated mass would closely correspond to the mass of the parent particle. Above the mass of the Higgs boson, nothing predicted in the Standard Model decays into two photons, so the photons essentially form random combinations in the absence of new particles. What physicists from both ATLAS and CMS saw was a tiny blip above the random combination background, at six times the mass of the Higgs boson, or about 750 GeV.

In the months following the December 15 announcement, CMS physicists did further analysis of the aspiring bump by using every bit of information they could find. First, they included approximately 20 percent more data than was used in the first analysis by including data recorded when the CMS magnet was turned off. Next, they applied a more precise calibration of measured photon energy, potentially reducing some of the uncertainty in the measurement. Finally, they combined data accumulated in the 8-TeV LHC run in 2012 with the newer 13-TeV data collected last year.

These U.S. physicists contributed to this analysis.

These U.S. physicists contributed to this analysis.

The results, which are presented in a paper accepted for publication in Physical Review Letters, show the December bump still there, rising 3.4 standard deviations above the predicted background at 750 GeV. Taking into account the full spectrum of masses searched, however, the significance of the bump is calculated to be 1.6 standard deviations, not much better than the odds of a coin flip.

As particle physics experiments search for rare events that are often dwarfed by background, bumps that are essentially statistical noise can come and go. The hope, of course, is for a bump that sticks around and becomes a new discovery, as the Higgs boson did in 2012. The frustration of the 2015 diphoton bump is that there simply wasn’t enough data at the time to determine whether it was due to noise or to a possible new discovery. Fortunately, it’s now more than halfway through 2016, and the LHC has delivered over four times the number of collisions as it did in 2015 to ATLAS and CMS. The analysis of these data should be completed in time for the International Conference on High Energy Physics, held in Chicago, in August. Expect to know then whether the diphoton bump has grown to be a clearer sign of a new particle or vanished the way of many bumps that came before it.