ATLAS/CMS Dark Matter Forum meeting

Europe/Zurich
4/3-006 - TH Conference Room (CERN)

4/3-006 - TH Conference Room

CERN

110
Show room on map
Antonio Boveia (CERN), Caterina Doglioni (Universite de Geneve (CH)), Sarah Malik (Imperial College Sci., Tech. & Med. (GB)), Stephen Mrenna (FERMILAB), Steve Mrenna (Fermi National Accelerator Lab. (US)), Steven Lowette (Vrije Universiteit Brussel (BE))

Notes on DM Forum Meeting, 28 January 2015

Agenda: https://indico.cern.ch/event/369001/

Upcoming meeting: https://indico.cern.ch/event/371630/ (Friday 6 February)

Open issues needing study are identified in italics.

Mono-jet models

Agreeing on a basic set of mono-jet like models was the focus of the
first part of the meeting. It was decided to start from these five models:

  • neutral s-channel mediators

    • vector
    • axial vector
    • scalar
    • pseudoscalar
  • colored t-channel mediators

    • scalar

In all cases the baseline dark matter particle will be a Dirac
fermion. The reasons for these choices were discussed at the meeting
and will be laid out in the document.

There was much discussion about potential choices for the couplings of
these mediator to DM and quarks. Of particular interest:

  • of 1) vector coupling to quarks and vector coupling to DM 2) axial
    vector coupling to quarks and dark matter 3) vector coupling to
    quarks and axial vector coupling to dark matter 4) axial vector
    coupling to quarks and vector coupling to dark matter, are 3 or 4
    distinct from the first two? This needs a generator-level study of
    the kinematics
    . The translation to direct detection experiments
    should be straightforward but hasn’t been done.

  • it was suggested to that MC samples for the mono-X analyses should
    use the ‘minimal’ width (given the quark and DM couplings, one can
    compute the total width of the mediator into quarks and DM). For
    mono-X searches, the width has only subleading effects on the
    kinematic distributions; the main effect of changing it is to adjust
    the production cross section. For direct mediator searches, such as
    bump searches, the minimal width should be augmented by another
    sample at a larger (but still physically interesting) width,
    representing additional decays into other particles. This choice
    should pay attention to the absolute width as well as relative
    changes in it. The choice of larger width requires further
    discussion.

    • the width of the scalar and pseudoscalar mediators requires
      particular care. there are two regimes: ‘dark matter’-dominated
      decay for mediator masses below the ttbar threshold and
      top-dominated decay above the threshold. 

    • the width of the vector mediator could also be treated as minimal,
      even though earlier studies and discussions show that the kinematics
      should not change as much as for the scalar and pseudoscalar. It would
      be good to make the relevant kinematic plots at generator-level for the document
      given that the models are available and the width can be changed
      (coherently with the couplings).
    • the width discussion is clearly tied to the range of quark and
      DM couplings we should consider. The Bristol workshop proposed a
      range of gq = gdm = 0.5 - 1.45 for the vector and axial-vector
      mediators, set on the low end by experimental sensitivity and on
      the high end by bounding the relative width at 100% of the
      mass. we need a a proposal for the scalar case as well (Uli,
      Valya?) All of these choices can be studied cheaply with
      comparisons of generator-level kinematic distributions, then
      used to inform the choice of a much smaller set of points for
      the experiments to simulate.

    • study of asymmetric couplings might also be considered. for
      example, one might fix the mediator mass and produce a scan in
      gdm vs gq. while the total production rate goes like the product
      of the couplings, dijet resonance searches may restrict gq to a
      greater degree than other searches restrict gdm. One should
      work out the constraints from dijets, ttbar, and other searches
      on these mediators
      . There are several people interested in
      doing this.

There have been two t-channel variants proposed, but these are
essentially the same. The main difference between these models and
SUSY squark mediation, other than the spin of the WIMP, is that the
coupling between the DM and the squark is kept as a free parameter
whereas it is fixed to a small value in SUSY. As a consequence, SUSY
searches can neglect some diagrams (squark neutralino production)
because it is small, whereas here that contribution can be very
important when gdm is order 1. It may be possible to reparameterize to
make use of SUSY samples in the small gdm limit, but new event
generation is needed for large gdm.

A colored vector t-channel mediator with vector DM could also be
interesting and needs more study.
 The UV completion of such models
would look like an RS KK gluon, for example. In these case, it is
expcted that the most interesting mediator masses will be of order
several TeV.

The possibility of a spin–2 mediator was raised (one of the few things
we know is that DM couples gravitationally). This has not been
studied.

Criticism of these choices (especially if you did not attend the
meeting) is more than welcome. It should be sent to
lhc-dmf-admin@cern.ch.

High priority follow-up:

  • Are the experimental signatures (kinematic differences) very similar
    for some subset of these models? The organizers again encouraged
    analyzers to perform comparisons of the kinematic distributions
    amongst the models and for a given model amongst different points in
    parameter space
    . This can be done with generator-level four
    vectors, to start. Some CMS analyzers and theorists (in particular
    the authors of http://arxiv.org/abs/1411.0535)
    have already done much of this and were encouraged to make their
    full suite of plots available to the forum.

Scalar model studies, Priscilla Pani, Kevin Kai Hong Sung

Priscilla and Kevin have implementations from Matt Buckley and have
begun the sorts of comparisons encouraged above for the mono-jet
models. They have begun to compile suggestions for where in parameter
space to scan coarsely and where to scan finely; what combinations of
parameters produce qualitatively different kinematic distributions;
etc. Some differences between the two studies need to be
understood. Priscilla and Kevin will continue their work and together
with others produce a concrete proposal.

Mono-top models and searches, Giacomo Cacciapaglia, and Romain Madar

Giacomo and Romain presented the current usage of mono-top models in
the experiments and discussed in what directions these should evolve.
A list of models and parameters to be tested for the mono-top signature
should be prepared in the same fashion as other signatures,
including the suggestion of further signatures sensitive to these models
(e..g same-sign dilepton) as in Romain's talk.

There are minutes attached to this event. Show them.
    • 16:00 17:00
      Discussion of monojet-like models
      • 16:00
        Guide for discussion 20m
        Speaker: Ms Caterina Doglioni (Universite de Geneve (CH))
        Slides
    • 17:00 18:20
      Discussion of top-flavored models
      • 17:00
        Summary of scalar models 20m
        Speaker: Priscilla Pani (Stockholm University (SE))
        Slides
      • 17:20
        Summary of ttbar+DM models 20m
        Speakers: Kevin Kai Hong Sung (Northwestern University (US)), Kristian Hahn (Northwestern University (US))
        Slides
      • 17:40
        Summary of mono-top models 20m
        Speaker: Giacomo Cacciapaglia (IPN Lyon)
        Slides
      • 18:00
        mono-top models and collider searches 20m
        Speaker: Romain Madar (Univ. Blaise Pascal Clermont-Fe. II (FR))
        Slides