Technically it also creates a diboson final state - two photons - but no, here I am not going to talk about the tentative new particle of which ATLAS and CMS continue to see hints in their data, at a mass of 750 GeV and with characteristics that increasingly resemble those of a heavy higgs boson. Oh, see - I am doing precisely that. It is admittedly hard not to speak of that thing nowadays, but I will insist, as I think it is too good to be true, and so it must be false. 

Today let me rather give you an update on some of the recent studies that the LHC experiments have performed in Run 2 data. There's heaps of them, and many of you would probably pick dozens of other topics as more exciting than the one I chose; but we have to start somewhere. So this time I will be discussing measurement based on events containing vector boson pairs.

Elementary vector bosons 

Although if we do not specify the word "elementary" the category includes heaps of composite hadrons of arguably smaller interest, what we particle physicists call "Vector Bosons" are the elementary particles that mediate the interactions between matter corpuscles. We are made of quarks and leptons -precisely, up and down quarks constitute atomic nuclei, and electrons make atoms with them- but these "matter" particles need forces to be kept together. The forces are provided by the exchange of photons, which transmit the electromagnetic interaction; of gluons, which mediate the strong interaction that binds quarks together; and of W and Z bosons, which mediate the only apparently less useful "weak interaction". So photons, W and Z bosons, and gluons are called "vector bosons" because of that crucial function. And they are, indeed, elementary, as far as we understand.

In many cases Gluons can be set aside from the above mix, as they do not "feel" the electroweak interactions. Their physics is quite different - they obey QCD, a tough theory we cannot completely compute-, and their study is complicated because experimentally they can be distinguished from quarks only on a statistical basis. You cannot look at any given event and say "look, here's a gluon": your chance of being wrong is at least 50%, while with photons or W or Z boson signals you usually get it right.

Also here I will stick with "vector" bosons, so the Higgs boson is also not belonging to the final states discussed in the remainder of this article. One reason for doing this is that the production of Higgs bosons is a rare phenomenon, so the statistical precision with which we can test the electroweak model with Higgs events is limited. Don't get me wrong: studying Higgs production is extremely interesting, but again, we can leave it for another day.

So finally if we say "electroweak diboson production" we may agree to consider WW, WZ, ZZ, Wgamma, Zgamma, and gamma-gamma final states. Six in total. But the situation is more complicated, as we have many ways to study the signature of each boson. Leaving aside photons, which do not decay (although they do convert to electron-positron pairs when they pass close to heavy nuclei, but we can omit this detail), the other two have many ways to decay, and therefore experimentalists have the choice of what signatures to go after.

A host of signatures

Let us take the W boson: it decays 67% of the time into a pair of hadronic jets. The other 33% are decays to electron-neutrino, muon-neutrino, or tauon-neutrino pairs (11% each). If you want to keep track of all of these different final states you will have your hands busy with quite different event topologies. 

And then take the Z: the Z decays to jet pairs 70% of the time, and 3% each to the three charged lepton-antilepton pairs. As for the 20% remaining decays, they produce pairs of neutrinos. If the event produced a lone Z at rest in the laboratory, there would be no way to spot its decay to a pair of neutrinos, as the latter cannot be detected directly - they escape the detector unseen. But if you look for a Z with large transverse momentum, e.g. one that is recoiling against some other energetic boson, then the neutrinos will indirectly be traceable due to the fact that the event will not "balance out" in the plane transverse to the beam: all the measured bodies flow in one direction, so conservation of momentum allows you to infer the presence of something which left without a trace.

One further complication is due to the fact that if the boson has a large momentum, the two jets it often decays into (take a W->jj decay for example) will "merge" into a single "fat jet". Not to worry: in the last decade ways to identify the two "sub-jets" in the fat jet have blossomed, and actually the reconstruction yields nice and clean boson signals; but the peculiarities of this topology further increase the pool of possibilities for dibosons.

To summarize, there are a host of ways to look for vector boson pairs in LHC data and thus test the electroweak calculations of the standard model. As you may imagine, basically all of them are being pursued by the sleepless analysts in CMS and ATLAS. The reason for this insistence is not only to check the standard model prediction. The calculation could be wrong, or in need of improvements; but more exciting is the fact that any deviation could signal the presence of new physics. And indeed, there were some reported excesses in Run 1 data. These called for more detailed checks with the new collisions.

Recent results

So let us see what the experiments are reporting. First of all, I should remind you that the 13 TeV collisions analyzed by ATLAS and CMS from the 2015 run of the LHC are not so many - a fifth of the data collected in Run 1. However, since the collisions are now at 60% more energy, massive final states are produced at higher rates. For the typical diboson searches this is not enough to make up for the data deficit, but one should keep in mind that the production of boson pairs by the decay of a new massive particle could be highly favoured (by one order of magnitude, that is) by the higher collision energy. 

Next, let us see a few selected plots, as I think I am writing too much. The first one is a graph of several measurements of the WZ production cross section, obtained by ATLAS looking for the very clean "trilepton" final states - you get them when a Z decays to electron or muon pairs, and the W decays to a electron-neutrino or a muon-neutrino pair.

The graph reports the ratio between the measurement and the standard model prediction. As you can see, there is an evident mismatch between the results and the expectation: all results are significantly above the theory calculation /so that the result for the measured ratio exceeds 1). Is this a hint that there is new physics, or rather, that the powheg calculation is missing some diagram ? I let you form your own opinion, but I note that theorists less than 2 years ago made a lot of fuss of a departure of WW production rates measured by ATLAS from theory predictions. Only later was it realized that the "jet veto" that the experimentalists were applying in their analysis was causing some unforeseen bias in the Monte Carlo prediction. Hence I prefer to suppose that here, too, the excitement will peter out.

[If you wonder what a "jet veto" is: as jets are the result of energetic quarks or gluons, and these particles are more often produced in background events than they are in diboson processes, experimentalists sometimes reject events that feature energetic jets when they want to select a clean sample of vector boson pairs. But in so doing they focus on a restricted region of the phase space of the studied process, and this may cause trouble, as mentioned above.]

The other two graphs I would like to show here are a summary of different measurements of several diboson final states, taken by CMS and ATLAS using data at 7, 8, and 13 TeV. Not all analyses have been finished, so there are some missing pieces to be added to the puzzle.  However here you can see that in general there is a good agreement between measurements and predictions. 

Above: the CMS measurements (empty red bullet are 7 TeV results, full red bullets are 8 TeV results, and full black bullets are 13 TeV results); below: the ATLAS measurements. In all cases the results are reported by dividing the measured cross section by the SM prediction, so that one expects to measure "1.0" if the SM is correct.

In summary, diboson production remains an interesting field of investigation, especially with the restart of the LHC taking place as we speak. Will the high WZ rates measured by ATLAS be shown to be due to some new particle decay ? I doubt it, especially after having seen the CMS result at 13 TeV, which is below the prediction...