Exotic Higgs Decays: Making the Case

For those of you curious about why posts have been a little sparse this month and have been wondering what I’ve been up to, here’s the latest on what has been an on-going story.

The Large Hadron Collider [LHC] is producing lots of new data, and the search for the Higgs particle continues at the ATLAS and CMS experiments. We’re still within Phase 1 of the search for the Higgs particle (I’ve described the two main phases here and in more detail here) in which the experiments are trying to discover unequivocally, or exclude unequivocally, the simplest possible form of the Higgs particle, which is called the Standard Model Higgs. Phase 1 is well along the way, the experimenters having excluded a Standard Model Higgs particle at any mass except a small range around 125 GeV/c². Within that range there are hints that a Higgs particle might be showing up (see here, here, here and here). Some would say the evidence is significant and are quite convinced already, while others (quite a few of the experimentalists, and a few cautious theorists such as myself) would say the hints are not yet especially significant and are willing to let more data settle the issue. But everyone agrees there’s a very high chance the issue will be settled in 2012.

At that point, the Higgs search will move toward Phase 2. In fact, in some sense it is already in transition, because either there is a Standard Model Higgs particle with a mass near 125 GeV/c², or the Higgs particle, if it exists, must be more exotic and complicated — two possibilities on which we can focus our planning.

One of the most important questions that will be asked (and is already being asked) is whether the particle that is showing up in the 2011 data (if it is really there in the first place) is a Standard Model Higgs or a look-alike Higgs particle that is actually more complicated in some way. This is a crisp and clear question, because the Standard Model (the equations we use to describe and predict the behavior of all the known particles and forces) predicts everything about the Standard Model Higgs except for its mass. Once its mass is known, all the rates for the various processes in which it is produced, and all the probabilities that it decays to this or that collection of other particles, are predicted, with a precision that varies between about 1% for the easiest calculations and about 20% for the hardest ones. If

• it is discovered that any of these rates are not equal to expectations, or
• it is discovered there are ways to produce the Higgs that are not expected, or
• it is discovered that the Higgs particles can sometimes decay in an unexpected (“exotic”) fashion,

then it will mean a revolution.  For we will then know that the Higgs particle and Higgs field are more complex than their Standard Model form, and that the Standard Model is not a complete set of equations and must be revised.  It is therefore critical to measure rates of production and decays of Higgs particles as well as possible, both for the processes that the Standard Model predicts will be present and for the ones that it predicts will be absent.

During the past few months (even going back to December 14th, right after the holiday Higgs hints were announced at CERN, the LHC’s laboratory) I’ve brought up several times (see here, here, here and here) the possibility that the Higgs particle (or particles) can sometimes decay in an exotic fashion, and emphasized that there are important but tricky opportunities to look for these decays in the coming year. In particular, with hundreds of thousands of Higgs particles expected to be produced in 2012, even very rare exotic decays might in some cases be discoverable. But there are two great challenges.

Analysis: First, a proton-proton collision at the LHC in which a Higgs particle is produced and decays is almost impossible to uniquely identify as such.   There are always other types of collisions, in which no Higgs particle is produced at all, that will mimic the decay of a Higgs. The Higgs decays  can only be picked out statistically, by gathering many similar-looking collisions, and determining that there are more than would be expected in the absence of Higgs particle production. We saw this in the holiday Higgs hints, where for instance the sign of a Higgs particle decaying to two photons was in the form of a bump on top of a smooth curve — an excess of collisions consistent with a Higgs decay over a “background” of two-photon events from non-Higgs processes. (You can read here for more discussion of this case.)  We also saw this last summer [the data in the article is out of date, but the issues that it addresses remain relevant], where there was an evanescent hint of a Higgs decaying to a lepton, anti-lepton, neutrino and anti-neutrino. So one must always ask: for a given decay mode of the Higgs, where a Higgs particle disintegrates into a particular collection of other particles, will there be enough Higgs decays (i.e. is there a large enough signal) and will similar-looking background processes be rare enough (i.e. is there a small enough background) so that these decays can be found if they are present, or if absent can be excluded meaningfully.

Triggering: Second, it isn’t obvious that these Higgs decays will even be stored for data analysis in the first place. The data from most collisions at the LHC have to be immediately discarded; there simply isn’t enough computer power available to analyze and store all of it. Fortunately, most proton-proton collisions at the LHC are boring: glancing blows, minor smash-ups, collisions in which no unknown phenomena occur. That’s why throwing most of the data away, using an intelligent automated algorithm called “the trigger”, is not crazy. But there is always the risk that the trigger strategy used to choose the one in a million collisions (i.e., the few hundred collisions per second) worth keeping will “throw the baby out with the bathwater”. The experimentalists worry about this a lot; theorists, in my view, ought to worry about it more than most of them do. The problem with exotic Higgs decays is that they lie in the danger zone. The particles produced in the decay of a Higgs with mass of 125 GeV/c² have relatively low energy and “transverse momentum” (i.e., momentum in the directions perpendicular to the directions of the incoming protons) — typically something near to or less than the 125 GeV or so of the Higgs mass-energy (i.e. E = m c² energy). Unfortunately, that much energy and momentum is produced in over a million boring collisions each second at ATLAS and CMS. From this huge reservoir of collision data, a trigger strategy must be found that winnows the number of boring collisions down by a factor of over ten thousand, yet still manages to keep a large fraction of the exotic Higgs decays, should they be present.  And whatever this strategy is, it has to be executed in much less than a second, typically using only partial and imperfect information about each collision.  It’s a tall order!

As a collider theorist familiar both with the theories (of which there are many) that predict exotic decays of various types (of which there are many), and also with the challenges of triggering (of which there are many), my highest priority over the past six months has been to persuade my colleagues at CMS and ATLAS that a program to search for exotic decays of the Higgs particle represents both a great opportunity for research using 2012 data and potentially a big problem for the standard trigger strategies. (I wrote about the opportunities and the potential problems back in January.) After some weeks of work on the subject, I summarized my concerns in a document that I sent around in mid-March to experts at ATLAS and CMS.

But at that point, although my studies indicated there might be a problem with the triggering strategy, I hadn’t had time to show that fixing the problem would actually have any real benefit as far as potential discoveries. In particular, it hadn’t been shown (and some leading members of ATLAS and CMS were not convinced that it could be shown) that the signals could be large enough, and the backgrounds small enough, that collecting these collisions using new triggering strategies would actually allow exotic decays to be discovered in 2012. And since setting up and testing new triggers is a huge amount of work, the skeptics did not want ATLAS or CMS members to be wasting time on what seemed to them to be potentially a fool’s errand.

This point of view came up in public during the SEARCH workshop, where I made a few remarks about triggering and exotic decays, and where some of my colleagues at ATLAS and CMS politely but firmly made their own skeptical positions clear.  We all agreed that the case needed to be made more convincing.

In response, a number of younger theorists who heard this discussion met during the SEARCH workshop, and a small group was informally established to address the objections of the skeptics. Is it true that more efficient triggering on exotic decays of Higgs particles could perhaps produce signals large enough, with backgrounds small enough, to allow potential discoveries in 2012?

These types of studies are not easy and take a lot of time and personnel, and all of the young theorists involved had other important things to do too. But conversely, we had no time to waste; if our concerns are well-founded, then unless they are dealt with right away, within a month from now, it will be too late: the majority of the 2012 data will be taken without the appropriate adjustments to the triggering strategy.  So we’ve been rushing, and we’ve started sending our results, piecemeal, to the ATLAS and CMS folks. We’re doing this as soon as our studies are roughly completed, once we have high confidence in them, but well before we’d be entirely comfortable submitting them to a journal for peer review… there’s just no time to do all the final checks that one would do under normal circumstances.

The good news is that I think our first studies give examples of cases for which the concerns of the skeptics are unwarranted. In particular, we think we’ve shown with reasonable confidence that searches for decays of the Higgs to two unusual clusters of particles would indeed become more sensitive if new trigger strategies were put in place.

More precisely (for experts), improved triggering on events with two unusual jet-like objects, with or without missing transverse momentum, appears likely to increase sensitivity significantly, relative to what would be possible were one to rely only on the “baseline” events from “associated production”, in which a Higgs is produced along with a W or Z particle, and in which triggering is based on any lepton or anti-lepton(s) produced in the decay of the W or Z.  Classes of “unusual jets”, and possible triggering strategies, are described here.  (Please feel free to contact me if you’re a member of ATLAS or CMS and want a copy of the current draft document. )

Our results are certainly not comprehensive, and we don’t expect such positive conclusions for all manner of exotic Higgs decays. But I think we have shown convincingly that the skeptics’ objections do not hold universally, and this should be enough to motivate some attention to the triggering issues.

My current collaborators in this effort: Tomer Volansky (at the Weizmann Institute), Jessie Shelton (Yale), Andrey Katz (Harvard), and (from Stony Brook) Rouven Essig, David Curtin, and Prerit Jaiswal, along with Raffaele Tito D’Agnolo (Scuola Normale Superiore, INFN Pisa, and CERN). Also, a number of our experimental colleagues at ATLAS and CMS have provided advice that has helped us ensure our studies actually are meaningful, sensible, and (we hope) valuable; thank you for your assistance!