Higgs excluded from 130 GeV to 480 GeV (Illustrative)

There are a few interesting workshops and conferences on today that are presenting results from LHC and Tevatron. In particular the “Implications of LHC results for TeV-scale physics” meeting at CERN all this week is the most likely place to look for new results, and indeed the following plot has just been shown by Eilam Gross.

This is an “illustrative” combination of the ATLAS and CMS Higgs searches which appears to be based on the data presented at lepton-Photon-2011. If you look carefully at where the black line crosses the 95% confidence level limit you will see that it excludes the standard model Higgs between 130 GeV and 480 GeV.

A Higgs below 130 GeV disfavours the standard model on its own because of vacuum instability. It might be OK if the vacuum remains metastable with a sufficiently long lifetime but if the mass is a bit smaller then such a universe becomes a very dangerous place to live. The safer explanation would be that the light Higgs is stabilised by extra particles which would have to look very much like a Higgsino or a stop. I.e. SUSY.

The excesses above 130 GeV are still there. It is difficult to read their size from this plot but they are obviously not due to a standard model Higgs. They could be from another boson with a smaller cross-section, or they may just be the effects of uncertainty in measuring the missing energy of the neutrinos in the WW channel.

No doubt more data will added soon and some possibilties are:

  • Excesses at 120 GeV could grow to a robust signal of a light Higgs, suggesting supersymmetry
  • The curve may continue to descend until the whole mass range is excluded according to the standard model
  • The 140 GeV excess could bounce back up from the grave to provide a standalone Higgs boson solution
  • The Higgs could appear at higher mass than 480 GeV, posing other problems for the standard model.
  • Some completely unexpected signal of electro-weak symmetry breaking could emerge.

Of course the plot is marked as “Illustrative” and I have no idea what other caveats the speaker has added (but see remarks in comment section from the speaker).

Update: For the record it turns out that the above plot used a combination formula which is not too good for observed CLs. See comments from its constructor below. It would be wrong to use it to draw any conclusions. You should think of it as an illustration of how missleading a combination can be if not done correctly. 🙂 It was removed from the uploaded slides.

My own combinations use a different formula which I believe is much better. They do not yet show an exclusion at 140 GeV.

22 Responses to Higgs excluded from 130 GeV to 480 GeV (Illustrative)

  1. Dilaton says:

    Dear Phil,

    thanks for continuing these nice reports about new or updated results from higgs and new physics searches (I still like them a lot) and for Your “emergency help” 😉 …


  2. I second Dilaton’s kudos for this excellent source of info.

    However, when I think of “new physics”, I am thinking of totally new ideas. For example: global discrete dilation invariance.

    I think nature is sending a clear message that the previous ideas for explaining the heuristic features of the standard model, and its shortcomings, have not worked out and that radical new ideas need to be explored without bias. That means questioning all inadequately tested assumptions.

    Discrete Scale Relativity

  3. JollyJoker says:

    Thursday’s “WG1: Constraints on the electroweak symmetry breaking sector from global fits” might give us some interesting news. Any bets on whether anyone is ready to call the SM dead already?

  4. Luboš Motl says:

    The Higgs below 130 GeV agrees also with the 114-128 GeV Higgs in the M-theory scenario by Gordon Kane et al.


    Also, another conference has detected – in a collision of two mosquitoes – the production of 50 TeV gravitinos, stringy moduli, squarks, sleptons, and heavy Higgses, proving that we live in 11-dimensional M-theory on a singular G2 holonomy manifold.

    The previous paragraph is illustrative. 🙂

  5. ohwilleke says:

    Odds are good that we get a resolution of this particular puzzle by New Year’s if not sooner. I’d be less surprised by a no Higgs result in this mass range than a light Higgs but I’m happy to see what we get.

    If we do get a light SM Higgs, the prospect of no new physics until post-LHC (if ever) looms large and it gets hard to justify a next LHC at all. If we don’t, the theorists have their work cut out for them, even though it shouldn’t have much of a predicted phenomological effect until we get closer to 1 TeV energy scales.

    • Luboš Motl says:

      Dear ohwilleke, you’re confused about both main points. A light Higgs boson (below 130 GeV) is

      1) more likely a priori – because of SUSY; because of high-precision pre-LHC determination of the Higgs mass (which should be low); and so on

      2) indicates that it is more likely that new physics *does* exist – some form of SUSY or something similar is needed to stabilize the Higgs potential when the Higgs is light.

      Note that you wrote the opposite thing on both points.

      • OXO says:

        And he also got this wrong Luboš:

        “the prospect of no new physics until post-LHC (if ever) looms large and it gets hard to justify a next LHC at all”

        Actually, it makes it easier if the machine you have is too small to make discoveries.

      • Luboš Motl says:

        Well, OXO, that’s an interesting strategy. 😉

        The next collider was largely supposed to be a lepton machine with energy reach not really higher than the LHC. It could only study similar phenomena as the LHC but more accurately. If there are no new phenomena at the LHC after 1,000/fb or so, except perhaps for a single 119 GeV Higgs, then there will be no rational justification to study these non-existent phenomena more accurately.

        One could build a similar collider with a higher energy but it could be a risk that it will see nothing, too. When people learn that there’s nothing between 140 GeV and 10 TeV, it will become likely – although not guaranteed – that there’s no physics between 10 TeV and 40 TeV (SSC center-of-mass energy), either.

      • Philip Gibbs says:

        If the LHC only finds a standard model Higgs up to 14 TeV then it will be very hard to get funding for another big collider. It may be possible to refurbish the LHC with stronger magnets to double the energy again up to 28 TeV centre of mass energy. There are already plans for that, but it will still be quite expensive, and may still not find anything new. That is not an easy sell in a budget strapped world.

        There is still plenty of scope for the LHC to find new physics though.

    • Lawrence B. Crowell says:

      Beyond upgrades of the LHC I think it is more likely we will have to get smart about doing particle physics with cosmic rays. Of course that will be difficult, for there is little control on particle energy. We also will return to the old fixed target method instead of colliders where we can work in the CM. However, the atmosphere is a sort of grand Cherenkov detector; well so long as we can loft with balloons a sufficient array of detectors. Other possible schemes might be used instead.

  6. carla says:

    If 5/fb is predicted to give a 2.7 sigma signal for a 115Gev Higgs, what is the prediction for a 120Gev Higgs?

  7. chris says:

    Thanks for your constant Higgs updates, they are highly appreciated!

    i think that the situation for our vacuum is not quite as bleak as you mentioned :-). first these stability bounds are from PT based RG arguments. as a lattice guy you should know to take them with a grain of salt. from arXiv:0710.3151 (fig. 4) you can read off the current state of the art. so a 135GeV Higgs is just barely too heavy to carry new physics up to the Planck scale. A 120 GeV Higgs would imply the SM to break down somewhere between 10^7-10^8 GeV – still a quite comfortable distance to go.

    but kudos to you for clearly spelling out that the current higgs exclusion region is far more damaging to the SM (as an effective theory up to the gravity scale) than for SUSY.

  8. […] in het zakje te doen en te speculeren over het Higgs boson en de supersymmetrie. :bron: Bron: Vixra. Noot:Over die vacuüm instabiliteit heb ik eerder al eens geschreven. [&#8617]Deel deze […]

  9. Eilam Gross says:

    Illustrative is ILLUSTRATIVE.
    I have shown the plot just as a teaser for the need to do an official combination which I support).
    But it is based on a nonsense combination which is not even supposed to work with Observed limits.
    So you should treat it as ILLUSTRATIVE and please avoid comparing it with any other combination, official or non official.
    The only thing we can trust now are the officila numbers coming from the two experiments;
    IT is true that based on these a Higgs Boson between 145-460 is less likely, but we will not know till we do the correct combination.
    The approximate formulas work better for expected limits. It is WRONG to use them for observed limits.
    Eilam Gross

    • Philip Gibbs says:

      Eilam, thanks for clarifying the status of the plot. I already have caveats to let people know that the plot is “illustrative” and have now added it to the title just to be sure. Your slides indicate an interesting talk and it’s a pity they did not webcast this meeting.

      I don’t know what combination formula you used but I am finding that the one I now use reproduces the official observed limits closely enough to be very useful. The effect of any approximations should also decrease as more data is added. This allows us to look at a wide variety of different combinations quickly to get a clearer picture of what is going on. I encourage you to do more of it despite what the combination group may say 🙂

  10. Eilam Gross says:

    Dear Philip
    Its not a matter of trial and error.
    I base my calculation on a formula from
    ( I am one of the four authors)
    If mu=sigma/sigma_SM
    and mu_hat is the MLE of mu which reflects the data,
    then you find that the upper limit is
    The 1.64 reflects the CL.
    To derive the expected limit you set mu_hat to its Asimov value, i.e. BG only
    This value is true for both ATLAS and CMS and for the combination.
    As a result you find the formula which you used 🙂
    i.e. mu_combined,expected=1/sqrt(1/mu_up.expected,CMS**2+1/mu_up,ATLAS,exp**2)
    BUT if it’s the observed you are at, each experiment has its own mu_hat and you cannot use the formula as is. Moreover, you would need to fit together both experiments results.
    The combined limit might even be worse than one of the individuals.
    That cannot happen with the expected.
    There the combined is always better.
    So if you are “lucky” and both experiments observe a similar fluctuation, the formula might do a less terrible job (for observed) but normally it is just not adequate.
    To put a long story short, you can approximate the expected limits with this formula, but you should patiently wait or the official combination in particular for the observed limit
    There is no other way.
    p.s. Are you at the conference? HOw did you grab my illustrative plot so fast?
    p.s.2 Theorists should only use such formulas for expectations. They should never combine the observed data. For this they need the likelihood functions and at the moment I do not see these available. But if a theorist want to test his theory and suggest channels, he should derive expectations, which are translated to sensitivities.

    • Philip Gibbs says:

      Thanks for the reference to your paper. I will study it to get some more background on the methods.

      The formula you state above is the one I originally used for both expected and observed but it did not work very well for the observed, just as you claim it shouldn’t. However, I later did a better anaylsis of the statistics and found a better formula. You can see it at http://blog.vixra.org/2011/08/06/better-higgs-combinations/ It is a very unsophisticated derivation compared to yours but this formulas has been pretty reliable.

      Even if it is not exact the errors should get smaller as the data accumulates and will always be far less than the statistical errors, so I think it is perfectly valid to use this technique, especially if the official combinations are not being made available very quickly.

      In answer to your p.s. I saw the slides when they came up on indico shortly after they became available. I am not at the conference.

      To answer your p.s.2 I refer you to the cheeky response at the end of this talk from today by Jose Ramon Espinosa Sedano http://indico.cern.ch/getFile.py/access?contribId=88&sessionId=19&resId=0&materialId=slides&confId=141983 🙂

  11. Eilam Gross says:

    I will have a look at your other formula as soon as I have the time (a bit busy now)

%d bloggers like this: