© Matt Strassler [December 11, 2013]

It’s been a while since I gave you a status update on the search for supersymmetry, one of several speculative ideas for what might lie beyond the known particles and forces. Specifically, supersymmetry is one option (the most popular and most reviled, perhaps, but hardly the only one) for what might resolve the so-called “naturalness” puzzle, tightly related to the “hierarchy problem” — Why is gravity so vastly weaker than the other forces? Why is the Higgs particle‘s mass so small compared to the mass of the lightest possible black hole?

Back in mid-2011, when the Large Hadron Collider [LHC] was still very young, my colleague John Conway announced on his own blog (Cosmic Variance) that supersymmetry (specifically, supersymmetry as a solution to the naturalness problem, which I will call “natural supersymmetry” here) was essentially ruled out by the data from the ATLAS and CMS experiments at the LHC. Well, it took three seconds to see that statement was wrong — and the fact that people are still looking for signs of natural supersymmetry even now shows just how wrong it was. Why is it so hard to rule out natural supersymmetry? Because there are an enormous number of variations on this theme … a huge range of variants of supersymmetry that can address the naturalness puzzle. To rule them all out is a huge task! And it’s not one that a few months of LHC data could accomplish, by a long shot.

By mid-2012, with five times as much data, and many more ATLAS and CMS searches through their data, the situation was a little more complicated. At that point, it took a bit longer — a few hours — to check that natural supersymmetry wasn’t ruled out by ATLAS and CMS results. Two things complicated the investigation. First, there were results from a very important measurement made by the LHCb experiment (though the LHCb experimenters managed to confuse the public by erroneously claiming, twice, to have ruled out natural supersymmetry, or at least to have “put it in the hospital” — one of the more meaningless statements about physics that I’ve ever heard.) The other was the hints, soon to be a discovery, of a Higgs particle with a mass of 125-126 GeV/c²; this turns out to be light enough to be consistent with natural supersymmetry, but just a bit too heavy to be consistent with its simplest variants. Still, it was far too early to claim anything definitive about natural supersymmetry, or about anything else, at that time.

What about nowadays? Well, what took a few minutes in 2011 and a few hours in 2012 now takes six months of hard work. The 2012 data collected at ATLAS and CMS was far more extensive than the 2011 data, and the experimenters at ATLAS and CMS have had far more time to dig through it. After all their efforts, it’s now a major enterprise for anyone to understand, in a clear-headed way, which variants of supersymmetry are ruled out and which are not. But it can be done. We’re now, finally, in a position to start drawing important, though still incomplete, conclusions about natural variants of supersymmetry.

As I said, supersymmetry comes in a bewildering array of variants, each of which makes somewhat different predictions for the LHC experiments. Is it possible, using existing searches by ATLAS and CMS, to make definitive (or nearly definitive) statements about natural supersymmetry, not specific to any one sub-class of variants, but applying to all (or a very larger class) of natural supersymmetric theories? This is the question my colleagues (Jared Evans, Yevgeny Kats, and David Shih) and I addressed in a recent paper.

The answer, we showed, is “yes”. It’s a qualified “yes”, because there are a few small loopholes, but I do emphasize the word “small”… which is a tremendous improvement over mid-2012, when you could drive trucks through the loopholes. Here’s the argument:

First: we assume we are dealing with a variant of supersymmetry that is natural, which requires that Higgsinos (superpartner particles of the Higgs particles [recall supersymmetry requires 5 types of Higgs particles]) with mass below about 400 GeV/c² or so. [This is somewhat conservative; most natural supersymmetric variants would have them quite a bit lighter.]

Second, we assume the gluino (superpartner particle of the gluon, and very easy to produce) is accessible, in the following sense: it has a mass below 1400 GeV/c², low enough that the LHC would have produced some of these particles during the 2011-2012 data taking.

Then we note the following: if and when gluinos are produced in the LHC’s proton-proton collisions, then, in almost every natural supersymmetric model with an accessible gluino, one or more of the following phenomena commonly results:

Missing Transverse Momentum: Striking signs that the collisions are producing both observable and undetectable particles, with the observed ones obviously recoiling against ones that were not observed. Top Quarks and Antiquarks: particles which are rather heavy (the top quark mass is 175 GeV/c²) and often decay to an electron or muon, an (undetected) neutrino, and a bottom quark [or their antiparticles.] Large Numbers of High-Energy Elementary Particles: (quarks, anti-quarks, gluons, leptons, anti-leptons, or photons); typical numbers are between 3 and 10 elementary particles per gluino, and thus 6 to 20 per proton-proton collision.

Finally, we point out that searches for each one of these experimental signatures have been carried out effectively, and with almost no assumptions, at both ATLAS and CMS. The searches for the first two signatures are so strong that it is virtually impossible for a natural supersymmetric model in which gluino decays produce top quarks and/or missing transverse momentum to have been missed, as long as the gluino mass is below 1000 GeV/c², and often as high as 1200 GeV/c². If instead gluino decays produce few top quarks and barely any missing transverse momentum, but still produce many quarks, anti-quarks and gluons, then the constraint on the gluino mass is a bit weaker, perhaps as low as 800 GeV/c² or so, but typically closer to 1000 GeV/c². [If leptons, anti-leptons or photons are abundantly produced too, then the constraints become stronger again.] We also pointed out ways that the searches for gluinos in this latter category might be improved for even better coverage.

This set of observations rules out the vast majority of natural supersymmetry variants with the gluino in the accessible mass range up to or near 1000 GeV/c². Only variants with heavier gluinos, or with gluino decays that avoid all three of the signatures mentioned above, or with unnaturally heavy Higgsinos, evade this statement. This result is summarized in the Figure below. Importantly, in contrast to many of the initial (i.e. 2011 era) searches for supersymmetry, which relied on the three key assumptions of what used to be the most popular versions of supersymmetry

in any process, the number of superpartners can only change by an even number; the lightest superpartner [which is stable, by assumption 1] is a superpartner of a particle we know (and therefore, to avoid conflict with other data, an undetectable neutralino or sneutrino); the superpartners that are affected by the strong nuclear force are significantly heavier than the other superpartners of known particles,

our results apply even if you drop any or all of these assumptions. We also do not assume that supersymmetry is “minimal” — i.e., that the only new particles to be discovered are the superpartners of known ones (plus the extra Higgs particles that supersymmetry requires.)

Where does this leave the search for natural supersymmetry? You might say it is about 3/4 done, roughly speaking. For natural supersymmetry variants that do not have a gluino that was accessible in 2011-2012, many searches have been carried out for other superpartner particles — but, as my colleagues Jared Evans and Yevgeny Kats have shown, one cannot say yet, in this case, that there is comprehensive coverage. For example, many searches for top squarks, the superpartner particles of top quarks, have been done, but each search has to make a certain assumption about how top squarks decay. There are clearly possibilities for top squark decay patterns for which no current search is yet effective. The same is true for Higgsinos and other similar superpartner particles.

Ruling out natural supersymmetry in a way that is nearly definitive, with very small loopholes of any sort, will not occur until after the LHC runs for a couple of years with the ~13 TeV proton-proton collisions that will start in 2015. But I think, by early 2017 if all goes well, we should be able to say that virtually all natural supersymmetry variants with a gluino mass below 1600-1800 GeV/c² are excluded (unless one is discovered, of course.) Limits on top squarks and on Higgsinos will also be much more powerful by then. And that will leave very little space for natural supersymmetry… little gaps will still remain, but the plausibility of the scenario will be very, very sharply reduced from where it is now, which is already greatly reduced from where it was before the LHC began running.

I should mention a couple of loopholes in our logic. The biggest loophole is that we assumed that there were no new long-lived particles, or other weird phenomena, in gluino decays. Such possibilities require an entirely different set of strategies, and are hard for theorists to study without knowing the details of how the detectors measure long-lived particles, a very complex subject indeed. For some types of long-lived particles, existing searches are very powerful; for other types, searches haven’t been attempted at all, so coverage is surely spotty. Second, and probably much less important, we assumed that the masses of the gluinos, Higgsinos and all other low-mass superpartner particles weren’t all squeezed together, within a few tens of GeV/c². Such scenarios require special treatment beyond what we covered; however it is not clear such a scenario is still possible within natural supersymmetry. Finally, one can in principle imagine gluino decays that are so complicated that they would confuse any existing search. Whether such decays could exist in practice requires more theoretical research. There are probably some other loopholes, but we believe they are rather small.

Another thing that’s important for non-experts to understand, and that is highlighted by our work, is that when the ATLAS or CMS experimenters say “we have done a search for X”, where X is some type of particle or phenomenon or speculative idea, that doesn’t mean the search is necessarily useless for Y, where Y is something altogether different from X. It’s important that the experimenters are looking for supersymmetry not only because supersymmetry might be there but also because the strategies they use may actually uncover something else. Conversely, sometimes searches for something with no connection to supersymmetry is secretly a useful way to look for certain variants of supersymmetry. [In short: even if one happens to believe strongly that X doesn’t exist, that doesn’t mean one should believe the search for X is a waste of time! One should not berate the experimenters for spending time “searching for supersymmetry” or “searching for extra dimensions” or whatever. This would be foolish, because the same searches are useful and needed for discovering or excluding many other things too.] For example: in our recent paper, we showed that a certain search for extra spatial dimensions (specifically, a search for microscopic, instantly evaporating black holes) is one of the most powerful studies for ruling out gluinos whose decays produce large numbers of elementary particles.

All in all, I’d like to think our results represent a step forward in understanding what ATLAS and CMS now tell us about particle physics, and in indicating how searches of future data ought to proceed. One possible lesson of our work is that in many situations, a small number of broad, comprehensive and powerful searches for generic phenomena can often trump a very large number of highly optimized but narrow searches for very specific phenomena. The latter leave far more holes, and, unless they specifically generate a discovery, are less useful than the former for drawing strong and general conclusions about nature.

