The Plot Of The Week - SUSY After The Higgs Discovery
    By Tommaso Dorigo | January 15th 2013 04:31 AM | 9 comments | Print | E-mail | Track Comments
    About Tommaso

    I am an experimental particle physicist working with the CMS experiment at CERN. In my spare time I play chess, abuse the piano, and aim my dobson...

    View Tommaso's Profile
    I may have been unattentive to recent papers on Supersymmetry, but I got the impression that during the last few months the majority of SUSY phenomenologists have been keen on speculating on the few apparently non-standard features of the recently discovered Higgs boson, as published last June (such as the alleged copious decay of photon pairs, or the dearth of decays to tau lepton pairs or b-quark pairs - note: all are mostly back to what the standard model predicts, after the analysis of more data following the discovery announcements), but rather improductive on the front of taking the new signal as a pivot point for new global fits.

    The matter is not trivial, although quite important. Inserting the new information in global fits is not trivial because if you want to do things correctly you need to take in account a dozen different rate and limit measurements (which are often moving targets given the pace of new results coming out of LHC and the Tevatron) and reconcile them with the predicted branching fractions that a quintet of non-standard model Higgs bosons may exhibit; with the added trouble that in case more than a single state lies close to 125 GeV they may have been unresolved by the current LHC data, and their branching fractions and relative rates need to be accounted together.

    [Incidentally, a true SUSY freak could take the two different mass fits of ATLAS for Higgs decays to photon pairs and Z boson pairs, which lie over two standard deviations apart, and try to interpret them with the existence of two different particles: that would in my humble opinion be too much, but please note that some of us do bet on the fact that these two signals are different resonant states...]

    And the matter is of course important because if you believe in Supersymmetry you might want to try and classify the newly found boson at 125 GeV as one of the three neutral states that the less complicated class of SUSY theories predict to exist: two h, H cp-even states, and a CP-odd state A. Or some mixture of those. If you do that, you immediately get some constraining power to add to the many direct searches for SUSY signatures in the tan(β) versus M_A plane. The exercise may thus reveal where to look for the other Higgs bosons, for instance !

    So, a new paper by Bechtle, Brein, Heinemeyer, Stal, Stefaniak, Weiglein and Williams describes exactly the above plan, carried out in a software called HiggsSignals, which will be released soon. As a representative exercise, the authors have taken information about experimental rates and limits of neutral Higgs decays in all measured final states by the experiments at the LHC and Tevatron, together with limits from charged Higgs signals, and produced a map of delta-chisquared (a statistical measure of compatibility between experimental results and theoretical predictions) in the tan(β) - A mass plane.

    The figure below shows what one should be looking for in the near future if one believes that there are four more Higgs particles there  to discover. In particular, by assuming that the 126-GeV Higgs boson is the heavier of the two neutral CP-even states, then the best-fit region highlighted in black shows that, if one accepts the authors' sample choice of some of the SUSY parameters (except β and M_A), one gets the hint that the A mass is 101 GeV, tan(β) is 6, and the light-Higgs h mass is 92.3 GeV.

    Together with contours of different delta-chisquare the figure shows excluded regions in the plane. These are computed using experimental information complementary to that used for the best-fit point: the upper limits on signal cross section in the various search channels. The window to minimal SUSY is indeed getting narrow! These graphs used to have a vertical scale extending from 0 to 100, and now we are down to ranges of tan(β) from 2 to 14...

    Also note that in this scenario the h particle, the lightest Higgs, escapes the LEP II search limits by acquiring reduced couplings to Z bosons (I recall that LEP II limits a standard model Higgs at 114.4 GeV, but that if one assumes that the production cross section of h in electron-positron collisions is smaller than what the standard model calculates, then LEP II could have missed that particle).

    I congratulate the authors for this nice new effort, and I look forward to more complete analyses of the SUSY parameter space which take into account all of the information we have amassed on the presence -well on the absence- of Supersymmetry in various parts of the over 100-dimensional space of SUSY theories!


    I am afraid that the best fit region claimed in that paper is probably already gone. Being just a proceeding I guess the authors underestimated the attention their draft would have drawn in the blogosphere and did not care to update the results they presented at the time of the workshop. The paper is a proceeding of a workshop held last October, before the Kyoto meeting in November were a more stringent limit in the plane m_a-tan(beta) from h->tau tau has been released (see CMS-PAS-HIG-12-050). Curiously, the authors acknowledge that in a footnote, but don't seem to care much and fail to mention that now the new lower bound on tan(beta) for m_A=100 is not 8 but 5. Combining this new upper bound with the lower bound tan(beta)>5 from charged Higgs searches that they show in their Fig.3 (see also 1204.2760) there does not seem to remain much space left.

    Thanks for the note Stefano. Strange - when I write proceedings I always write about the situation at the time of writing, and not the one at the time of my presentation!

    As one of the authors, I would like to give a brief reply to this:

    It is true that there are more recent limits from CMS, limits which we are currently busy implementing into HiggsBounds. As already guessed by Tomaso, the proceedings are written using the information (and plots) available for the talk in October. I guess none of us expected the blogs to pick this up ;-)
    The specific parameter setting that was chosen (to illustrate the constraints from HiggsBounds+HiggsSignals) was actually the same as first discussed (by some of us) in [1112.3026]. The main point of the proceedings article was not to perform a full analysis of this scenario, but rather to present the result from the two codes on something "familiar".

    On the other hand, an important point about the limit you are referring to (in CMS-PAS-HIG-12-050) is that the published limit applies in the mhmax scenario only (!) For scenarios which are not at all similar to mhmax (of which the "heavy Higgs 126 interpretation" is an extreme example), direct application of this limit will not give you the correct answer. That said, it is clear that the new CMS limit will give stronger constraints also on this scenario (although not as strong as tan(beta)>5). This is something we are currently investigating in more detail.

    Right, it is always a delicate issue to carry exclusion plots
    obtained with particular assumptions on to other scenarios, so in
    order to draw robust conclusions a comprehensive analysis such as
    the one which you are now carrying on is required.
    If your proceeding was just the appetizer, let's wait for the main course!

    What's the ultimate integrated luminosity recorded by CMS and ATLAS at the LHC from pp collisions at a centre-of-mass energy of 8 Tev?
    Da Liu

    Thank you very much.
    Da Liu

    While I agree that the choice of data is confusing at the very least, I agree that this sort of effort is very welcome. Let me comment on two points you made:

    - "as published last June": for the chi^2 scan it seems they used the ICHEP data, from July. the citation for Higgs searches that exclude the top-left part of the tan(beta)-m_A plot you posted is Ref. 11 in the preprint. That is the ATLAS combination from February 2012, well before June, well before the July discovery announcement. Why the cocktail is beyond me.

    - "all [non-standard features] are mostly back to what the standard model predicts": ATLAS's diphoton results from the December 2012 update [ATLAS-CONF-2012-168] are still 1.8 time the SMH, consistent among production modes (albeit with larger uncertainties in the rarer ones).

    Hello Andre,

    sure, ICHEP is July, not June - but I can argue that saying June is not incorrect, since most of these results got approved then (although showed at ICHEP indeed). As for what goes in exactly, I also do not know the details and their rationale, but the focus of the post is elsewhere...
    About the last issue, let me just say "no comment" ;-)