The SUSY-Inspiring LHC WW Excess May Be Due To Theoretical Errors
    By Tommaso Dorigo | July 18th 2014 03:57 AM | 7 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
    A timely article discussing the hot topic of the production rate of pairs of vector bosons in proton-proton collisions has appeared on the Cornell arxiv yesterday. As you might know, both the ATLAS and CMS collaborations, who study the 8-TeV (and soon 13-TeV) proton-proton collisions delivered by the Large Hadron Collider at CERN, have recently reported an excess of events with two W bosons. The matter is discussed in a recent article here.

    Those departures of WW cross sections from theory predictions are not extreme - one could well explain them away as a statistical fluctuation, or the combination of a fluctuation with some common shortcoming of the two measurements; but they are interesting, as it would be exactly the kind of effect one would observe if a not-too-strange realization of Supersymmetry were what Nature has chosen for our Universe.

    Now Preirit Jaiswal and Takemichi Okui, from Florida State university, argue that the calculations of cross sections of WW pairs needed to compare to LHC data have to account for the experimental selections, which include a "jet veto". That is, both ATLAS and CMS to enhance their WW signals and reduce contaminations from competing processes (like top quark pair production) impose to the data a cut on the presence of hard jets in the events.

    A jet is a stream of particles emitted when a energetic quark or gluon is kicked off the hard collision region. The parton (that's what we call a hadron constituent, i.e. either a quark or a gluon) "dresses up" as a stream of particles, which collectively share the total energy and momentum of their common parent. Hadronic jets are a quite common observation in proton-proton collisions -it is quite hard to observe events with no jets! But indeed, requiring the lack of energetic jets in WW events is a very useful tool to enhance the signal to noise ratio. So where is the problem ?

    The problem is that ATLAS and CMS resort to Monte Carlo simulations of WW production with a parton shower modeling to see how many WW+0 jets events would pass their selections. These are quite well-studied and advanced tools, but they might fail to produce accurate estimates for the corner of phase space that is used to compare data and predictions.

    In the theoretical calculation of the process one has to deal with two different energy scales: the scale of production of the WW pair, and the scale at which a parton is radiated off by the hard subprocess. If you want to account for the presence of this parton, which becomes an observable jet, and calculate only the rate of WW pairs which do not feature any such jets, you need to operate some complex calculations called "resummation" of large logarithms of the ratio of the two energy scales.

    The authors employ a technique called SCET, for "soft collinear effective theory", which is way too complex to explain here (and to tell the truth I can't explain something I do not fully understand myself!). However, their paper is quite detailed and, I must say, exceptionally clear, so if you are willing to give a try at understanding how these calculations are done, the paper is a good read for you. Let us see what the authors conclude for the cross section of jet-vetoed WW boson production processes.

    The comparison of the authors' resummed NNLL+NLO calculations to ATLAS and CMS experimental results for the WW cross sections is shown in the graphs on the right; the top graph is for the 7-TeV measurements and the bottom one for the 8-TeV one. There is overall excellent agreement, with only one of the four results being just over one sigma away! The 2- and 3-sigma effects reported by the collaborations are no longer present. I believe this is a very good example of how two "excesses" seen by different experiments do not count as twice as much evidence for something new: indeed, both results may be affected by one and the same systematic error source.

    I believe that if one has to choose between the hypothesis that the excesses are due to Supersymmetry and the hypothesis that the jet veto introduces a large systematic uncertainty, it would be hard to pick the first. So, sorry SUSY enthusiasts - let's keep searching.


    When all findings of the LHC are affected by such quite obscure doubt raising criteria, then the LHC is loosing much of its confidence by non-expert readers.What kind of expert must the observer be in order to look through the rough results in order to judge what a clean trustworthy result will be?
    If you think, think twice
    Dear Hans,

    experts are always sceptical of any result not in line with expectations, for that very reason. Nobody can "look through the rough results" and see which are the trustworthy ones. Science progresses by trial and error... Get it over with !

    Well, as I ranted in the comments to your last post, who would really have believed that SUSY not only exists in nature (so far so good), but also tuned to weird unnatural-looking mass degeneracies between particles which are just right to elude early observation. It would have been a nice riddle for theorists, though! To cook up a susy model where one squark and the chargino are naturally both very light and degenerate up to a few GeV...

    Hi T -- that is great that the excess has a resolution. Thanks for letting us know.

    Dear T,
    I'm glad you found my paper interesting.
    However, I would not call it a 'theoretical error' since MC ain't expected to work perfectly in the region of the phase space used by the experimentalists for the WW measurement to begin with.

    Hello Prerit,

    sure, I know. Titles have to catch people's imagination, so I allow myself a little bit of "sensationalism" there. I hope the body of the blog does not contain blatant mistakes or misrepresentations.

    Keep up with the good work,
    This same group has updated and presented their results at SUSY 14'.

    Of particular interest is slide 18 where they acknowledge the paper cited in your article (arXiv:1407.4537) as well as a second paper (arXiv:1407.4481) regarding NNLL transverse momentum re-summation.