Vladimir Litvin, Harvey Newman, Sergei Shevchenko, Tony Lee

1
Search for graviton in ADD and RS-1 models
decaying into two photons

• Vladimir Litvin, Harvey Newman, Sergei
  Shevchenko, Tony Lee
• Caltech CMS
• Jim Branson, Marco Pieri
• UCSD CMS
• Marie-Claude Lemaire
• SACLAY
• Mikhail Dubinin
• MSU

2
Introduction

• Current status
• Request for Physics TDR

3
Status

• Preselection parameters have been confirmed for C
  lt 0.2 area and MG gt 1 TeV for the RS-1 model
• Minv gt 900 GeV
• ????? lt 2.4
• E??? gt 150 GeV
• E??? gt 100 GeV
• Pthat gt 100 GeV
• 70k for 10 fb-1 of selected events have been
  produced by PYTHIA and ready to go into
  OSCAR/ORCA production.
• ADD model was implemented in CMKIN and first
  results already been obtained. We can use the
  same background for the ADD model as well as for
  the RS-1 due to already existed astrophysics
  model parameters limitations (to be presented on
  the next CMS/CPT week)
• We shall make cross-checks with other generators
  in the second half of this year. CompHEP will be
  used for the comparison.

4
Status

• Pentium4 3GHz
• 1.2 hours per event
• 336000 generated events per one selected event
• Taking into account that QCD background
  cross-section for Pthat gt 100 GeV is 1.4 109
  fb, we need to produce 40k-50k of selected
  events for 10 fb-1. We need roughly 50k-60k
  CPU hours on P4 3GHz to generate these events.
  70K of selected events already been generated so
  far, 85k CPU hours on P4 3GHz has been spent.
• For 100 fb-1 we need 450k of selected events
  and 500k-600k CPU hours (P4 3GHz)

5
Request

• Q what is the topology under study
• A diphoton graviton decays in RS-1 and ADD
  models
• Q which signal generator is used (any problems?)
• A RS-1 PYTHIA. The angular distributions are
  not implemented correctly for the diphoton
  channel, even in the latest version of PYTHIA -
  6.225. (checked by Marie-Claude)
• ADD - STAGEN collection of generators for CMKIN
  3.1.0 - ADD model http//cmsdoc.cern.ch/cms/gener
  ators/www/geners/collection/stagen/stagen.html

6
Request

• Q what are the specific problems of this
  analysis related to OSCAR, FAMOS, ORCA and which
  developments of this software will you contribute
  to
• A There are three open issues so far
• electronic saturation, which was reported on last
  CMS week (CMS AN03/010 from Caroline Collard and
  M.-C. Lemaire, presented last CPT week by
  Caroline for the electron pair outgoing channel).
• potentially large leakage from the rear side of
  the crystals due to the large energy deposition
  in a single crystal
• The uncertainty related to the different proton
  parton distribution functions and due to the
  internal uncertainties in PDFs. This is unclear
  now how we can check that and take into account.
  (For the Drell-Yan, see Shmatov et al talk on CMS
  week - http//agenda.cern.ch/askArchive.php?basea
  gendacatega041651ida041651s7t34/transparencies

7
Request

• Q which experimental results do you intend to
  provide
• A Structure of the results will be very close to
  our Analysis Note 03/009. This is almost the same
  topology (of course much less energetic) and
  smaller number of backgrounds (one instead of
  four in CMS-AN03/009)

8
Request

• Preliminary list of figures and tables
• table with total cross sections for different
  graviton masses
• table with QCD background cross sections after
  our PYTHIA preselection
• number of generated and simulated events
• description of energy and mass reconstruction
• photon impact point correction function
  (Emeas-Etrue/Emeas vs S1/S9) as well as
  correction in function of the pseudorapidity eta
  as it was done for the ee- channel in
  aforementioned note
• invariant mass reconstruction for specific masses
• reconstructed mass with different corrections
  depending on our algorithm for the vertex
  renconstruction. The best will be chosen to
  produce final results.
• tracker isolation and ECAL isolation - table
  about how many events passed through each type of
  isolation and the final result - after all
  isolations
• invariant mass distribution for the background

9
Request

• Preliminary list of figures and tables (Cont'd)
• remaining background at different diphoton
  invariant masses
• table of signal efficiency, background
  cross-section and integrated luminosity needed
  for 5sigma discovery for different mass windows
  and different graviton masses
• the same table when K-factors will be taken into
  account. K-factor for the background will be K1,
  and will use K1 for the signal unless new
  results will come up.

10
Request

• Q which are your most relevant background
  contributions, both from Standard Model processes
  and from new physics
• A Standard Model QCD background is the most
  important. Nothing significant known from the BSM
  but we are investigating it... Nothing known yet
  for the NLO of QCD background. We will use K1
  for the meantime.

11
Request

• Q for which parameter values do you need (or
  will you make) full detector simulation
• A Preselection parameters
• Looking for events with at least one pair of any
  two of the following particles,
  ?0??e???????????????, satisfying
• Minv gt 900 GeV
• Pthat gt 100 GeV
• E??? gt 150 GeV
• E??? gt 100 GeV
• ????? lt 2.4

12
Request

• We will use default OSCAR parameters and default
  ORCA parameters. MGPA saturation problem will be
  fixed when reconstruction will be finished and
  digis will be produced. We have to determine the
  correlation of S9-S4 variable with S1 in the
  region where there is no saturation to be able to
  correct S1 in the region where saturation does
  exist, based on S9-S4 vs S1 parameter values.
• Another changes are possible later on. Will
  depend on our analysis

13
Request

• Q which test statistic do you use for evidence
  of a discovery
• A CMS approved test statistic will be used
• Q which systematics will be included
• A Uncertainties of PDFs
• Q will you (also) study the CMS reach (with
  FAMOS?)
• A No
• Q will you reconstruct new particle masses
• A Yes

14
Request

• Q which are the Institutes and who are the
  people involved in this analysis
• A
• Caltech Harvey Newman, Vladimir Litvine, Tony
  Lee, Sergei Shevchenko
• SACLAY Marie-Claude Lemaire
• UCSD Jim Branson, Marco Pieri
• MSU Mikhail Dubinin

15
Request

• Q is this manpower sufficient to meet the
  schedule of the Physics TDR
• A Basically yes, but it would be helpful to
  have a theorist who will check the generators and
  help us to fix the problems like recent problem
  with incorrect photons angular distribution in
  RS-1 model