Wednesday, June 27, 2012

Some totally healthy SUSY models

Matthew W. Cahill-Rowley, JoAnne L. Hewett, Ahmed Ismail, and Thomas G. Rizzo have looked at millions of phenomenological MSSM models (to be explained later) and they have found some "totally ordinarily looking" models which are not fine-tuned, which have light superpartners, and which are compatible with the 125 GeV Higgs boson as well as all published constraints on SUSY:
The Higgs Sector and Fine-Tuning in the pMSSM
Just to be sure, our ignorance about the precise way how SUSY is broken in the MSSM may be quantified by 100+ parameters. There is a subspace of this parameter space that has 19-20 parameters, the phenomenological MSSM or pMSSM, in which we require no new sources of CP violation, minimal flavor violation, degenerate 1st and 2nd generation squarks and sleptons, and vanishing A-terms and Yukawa couplings for the first two generations.

The LSP is the gravitino or a neutralino in these models. Incidentally, I've been thinking about the possibility that the LSP is charged, a stau, and we observe lots of "very heavy" hydrogen atoms (with the same spectrum) that have a stau or antistau in the nucleus which makes them hundreds of times heavier than the ordinary hydrogen atoms. Chemically, they're indistinguishable but their extra mass could replace dark matter, couldn't it? The main problem is that one would probably get too much of this stuff as the staus wouldn't be good enough in annihilation...

At any rate, starting from page 36, they showed the superpartner spectra of models that are not only compatible with all published measurements including the 125 GeV Higgs boson but that have also an unusually low amount of fine-tuning, \(\Delta\lt 100\). Here is the first one they show:

Click to zoom in.

They offer 12 other charts like that – and each of these models looks substantially different from others. Note that many superpartner masses are well below 1 TeV in these models.

In particular, the neutralino masses are near 100, 150, 250, and 3000 GeV: only the heaviest one is heavy. Both charginos are light, 100 and 250 GeV or so. The lighter stop and sbottom squarks are near 400 and 350 GeV, respectively. Their heavier cousins are near 1100 and 1600 GeV. Five slepton masses are also between 400 and 800 GeV while the gluino and the remaining squarks and sleptons are between 2400 and 3600 GeV.

The other 11 spectra they show differ in many important details. These models are compatible with the current LHC data, despite the light superpartners, for various reasons: generally, they just don't predict too much missing transverse energy, typically because a superpartner prefers to decay to other but heavier superpartners than the LSP. This produces chains instead of lots of missing transverse energy.

The last sentence of the paper brings an upbeat message:
Hopefully the LHC will discover both the light Higgs boson as well as the 3rd generation superpartners during its 8 TeV run.
Not bad because the first condition is likely to be satisfied in 7 days from now. What about the other one? Does the democracy between these two hopes suggest that they're aware of a 3rd generation rumor that is comparably strong to the widely known July 4th Higgs rumor?

Of course, if you believe that one of these models is right, you may ask: Why would the correct point in the MSSM moduli space try to imitate the non-supersymmetric Standard Model as well as it apparently does? In other words, why would the right supersymmetric model be one of those that require the longest amount of time to be discovered by the LHC – which seems to be the case, kind of? That's a good question but I would like to emphasize that it is a question. There is no proof that there can't exist a good, intuitive, qualitative answer. There may very well be a good reason.

If you agree that a phenomenologist should try to reduce the amount of fine-tuning, each of these models is much more acceptable than the Standard Model whose fine-tuning is extreme, despite the fact that one has to choose "the best model among 100,000 candidates" (which is not too much). On the other hand, if you think that this whole \(\Delta\)-counting and naturalness arguments are wrong and you prefer minimality, the Standard Model is your golden standard.

If the LHC continues to find no new physics, the latter position – the Standard Model – will be getting stronger relatively to the first one (if nothing else happens that would change the odds, of course). However, this drift is extremely slow. Whenever you reduce the surviving fraction of the MSSM parameter space by another factor of 2, you are effectively finding something like new 1-sigma evidence against SUSY. However, at the same time, people may use completely different theoretical or experimental methods to find some new 2-sigma evidence in favor of SUSY. So the exclusions of mild "majorities" of spaces isn't too strong an argument.

BTW if you haven't had a breakfast yet, it's very important for you to prepare it in a politically correct and mathematically correct way. Concerning the second condition, you have to cut the bagel in such a way that it is transformed to a pair of mutually linked tori. ;-) Instructions may be found in the video above. The essence of the procedure is familiar to those who have cut the Möbius strip in the middle and repeated the same process with the resulting non-Möbius, doubly twisted strip.


  1. This spectrum is pretty tuned. The gluino mass feeds into the stop mass at one loop, so even if SUSY is broken at very low scales, like around 10 TeV (a bad idea for other reasons) the one-loop RG estimate of the stop mass is going to come out at near 2 TeV. Then they need to tune the stop soft mass to keep the stop light. Tuning happens for more than just the Higgs! The gluino also filters into the Higgs at 2 loops, though.

    It's really pretty much a fact that any MSSM spectrum consistent with current data is tuned at the part-in-a-few-hundred level or worse, usually much worse.

  2. Hi, thanks, but I have a problem with this kind of a complaint. There are many different masses, many different parameters, so even if all quantities are natural and random, it's pretty much guaranteed by statistics that *some* masses will be 10 times lighter than the naive esimate, right?

    For this reason, I wouldn't dare to attack a model that would be a part-in-a-few-hundred fine-tuned according to your very inclusive rule: this mild tuning is almost bound to happen.

  3. I'm confused -- if dark matter is charged, would it not scatter photons? Or are you saying it won't because it's dark hydrogen and not dark protons + electrons?

  4. It will scatter photons but the amount of this scattering may be very low because the dark matter is heavy.

    If (anti)stau weighs 300 GeV, it will combine with electrons to produce heavy hydrogen atoms, 300 times heavier than the normal one. They will have spectrum just like the hydrogen atom - tritium has the same spectrum, too. But you will need 300 times lower an amount of this stau-hydrogen dark matter to induce the same gravitational effects as one hydrogen can.

    The question is how quickly the antistaus will annihilate against the electrons; and/or pair-annihilate with each other which would need to produce some charged products along the way. Those things have to be compatible with a cosmological model.

  5. You're probably right that they have a large A-term for mixing, but then that contributes directly to the Higgs tuning cost. I'm pretty sure the joint tuning measure considering both the Higgs and the stop will be significantly worse than part in a hundred.

    I'm not sure I buy your "look elsewhere" argument; some mass could come out unexpectedly low, but here it's two masses that are important for naturalness -- if the stau, rather than stop, came out much lighter than expected, it wouldn't help with the Higgs.

    Quantifying naturalness is not a very interesting thing to do, in the end; such spectra could very well describe the real world. But they are in an awkward place where neither true believers in naturalness or anthropicists can feel vindicated...

  6. Good comments. Nevertheless, concerning your last sentence: I, for one, have absolutely no problem with the possibility that Nature chooses a middle ground, vindicating none of the two groups. It's not a Nature's duty to vindicate one of the ideological positions. And I wouldn't consider such points awkward; they may be the ultimate Goldilocks points.