The ICHEP conference in Chicago is drawing to a close, and although I did not have the pleasure to attend it (I was busy with real work, you know 😉 I think I can post here some commentary of a few things I find interesting among the multitude of analyses and searches that were shown there. It goes without saying that the selection is biased by my personal interest, plus by my limited patience with peeking at talk slides. In fact, here I only cover one specific Higgs boson decay mode!
But a digression first – and a digression on the digression
Yet one thing I will not fail to mention for a starter is that the 750 GeV resonance is indeed gone from LHC data. Whether this is a good-bye or not will depend on how gullible we can prove ourselves to be: the chance that some more data piles up there to make a new significant fluctuation is larger than 50%, as previous data never gets erased.
Incidentally, here’s a fun fact to ponder on: if you keep looking at some effect as you go on collecting data, you are likely to see something significant at some point of time. You might then be tempted to stop data collection there and then, and proceed to publish your “excess”. This is called “sampling to a foregone conclusion” and is a different kind of the by now well-known Look Elsewhere Effect. One might conclude that the ways an experimentalist may fool him or herself are indeed many, and scientific rigor is never enough.
Anyway I will not even bother to paste here the money plots for ATLAS and CMS. In case you have been away during the past ten days or so, please check your twitter or facebook columns and you’ll find them in multiple copies.
The Higgs decay to bottom quark pairs
So now let’s move to more serious topics. One thing I really want to see in the new data is a first clear signal of H→bb decays – a personal bias due to having worked with heavy resonances decaying into final states with b-jets in my early career (the first all-hadronic top signal, and then the Z→bb process, both in CDF data).
As you might recall, the Higgs boson couples to fermions with a strength proportional to the fermion mass. This makes the H→bb decay the most probable one with a predicted branching fraction of 58% (the decay H→tt, where by tt I mean a top-antitop quark pair, is forbidden, because the Higgs boson weighs much less than two tops).
Indeed, by fitting together all Run 1 data, ATLAS and CMS could already extract information on the bbH vertex. This tells us that, yes, the coupling of the Higgs to bottom quarks should more or less be in line with what we expect. But a bump is a bump, and until I see a bump I don’t believe nothin’… Also when I see one, as some of you I think by now reckon!
So let me show some of what was seen in searches involving the decay H→bb. CMS for example sought for a process whereby the Higgs boson is produced together with two forward-going jets. This is called “vector-boson fusion” (VBF) and is a very peculiar process. Indeed, when you collide protons you don’t expect that the scene is stolen by a pair of W bosons, but that’s precisely what happens in VBF Higgs production: each proton emits a W boson, and it is those two particles that actually collide, creating a Higgs.
The top graph shows a mass distribution for b-tagged jet pairs. A Higgs boson signal could be seen there only if there was a large excess over SM predictions, but that’s not the case – indeed CMS sees a deficit at the expected mass. But give it time, and we’ll see a signal there one day!
ATLAS also made many different searches sensitive to the H→bb decay. Those where the Higgs boson is produced in association with a W or Z boson (which provides good triggering leptons, allowing an efficient collection of the events) returned a signal which is however much smaller than what the SM predicts. But this is of course only due to the fact that the data is as of yet insufficient for a definitive measurement of the processes. You can see the ATLAS results below, where the combined signal strength of the WH and ZH channels is found at 0.2±0.5 (i.e., fully compatible with zero, or with 1, or with anything in between).
A new rare channel thrown in the mix
Instead, I was very pleased to see ATLAS show results of a search in another channel, a rare one which I had considered many years ago with a PhD student. In 2007 we decided that it was not worth spending our time on it, because of its rarity in the SM. The process is a vector boson fusion-induced production of a Higgs boson and an energetic photon. The Higgs can then be sought in decays to bottom-antibottom pairs and the photon provides additional triggering possibilities with respect to those offered by the forward jets. The production diagram is shown on the right and the preliminary result of the search below.
Like CMS in its VBF H→bb search, ATLAS gets a best-fit signal strength which is negative, using the selected data shown above, and sets a limit at 4 times the SM expected production rate. As you can see in the graph, the signal might indeed become visible if we collect O(10) times more data (but don’t be deceived – the signal is scaled by x10, but if you take 10 times more data the background moves up as well…). Anyway, I find this a cool search and I will love it when we will eventually see a signal there!
ttH, with H→bb – a tough beast to tame
Finally, another analysis I entertained myself with a long time ago (when we had no data yet!) is the search for the associated production of top quark pairs and a Higgs boson, with the Higgs boson decaying into a pair of bottom quarks. This search is very complicated because of the large QCD backgrounds (mostly top pairs plus b quark pairs).
ATLAS did very well and they are now seeing some small signal there, as shown in the distribution of their multivariate discriminant (see below). Their fit gives a signal strength equal to 2±1 times the SM prediction, so we are starting to measure this process even in the difficult H→bb final state (other decay modes of the Higgs are also exploited for the ttH search, but let’s not discuss them here).
In conclusion, the H→bb signal is still not conclusively observed by the LHC experiments, but it is only a matter of time. And of course, dedication! There are in fact dozens of different search channels and it is their combination that will give the first branching fraction measurement that can stand on its own feet. So stay tuned if you are interested in this process, as 2016 data is continuing to accumulate. I think this will get very interesting by the time we get to the 2017 winter conferences!