Possibility of tan(b) measurement with in CMS

1
Possibility of tan(b) measurement with
in CMS
Majid Hashemi CERN, CMS IPM,Tehran,Iran
QCD and Hadronic Interactions, 12-19 March 2005,
La Thuile, Italy
2
Motivation

• tan(b) is one of the most important parameters in
  MSSM,
• It enters in all parts of the theory,
• Value of tan(b) can be measured at LHC in
  different ways depending on its real value,
• At large tan(b) (gt10) there is a possibility to
  measure this quantity with the Higgs sector using
  event rates, in this study .

NLO
In MSSM at large tan(b) ,
cross section exceeds that of
for almost all the mass values.



M. Hashemi CERN,CMS (IPM, Iran)
3
Cross section and branching ratios for the signal
events
At large tan(b) dominant parts of the total cross
section are proportional to which I
call tan(b) hereafter. Branching ratio
is almost constant as a function of
tan(b), keeping tan(b) dependence of the
cross section, tan(b) measurement
uncertainty half of uncertainty of rate
measurement.
PPHTTHDECAY
SUSY parameters are chosen according to the LEP
benchmark scenario
M. Hashemi CERN,CMS (IPM, Iran)
4
Signal and background identification and
simulation tools
Events are generated by pythia6,toprex,tauola ,
signal cross section with PPHTT and Higgs
branching ratios with HDECAY, Studies with
parameterized fast simulation, partly with full
simulation. Signal events are
Main Backgrounds are
Background events are rejected by Lepton
isolation, Tau-jet identification, Tau impact
parameter, b-tagging, Jet veto.
M. Hashemi CERN,CMS (IPM, Iran)
5
Signal event selection
Event trigger is done with L1HLT package for
each channel
Single muon trigger, eff.0.85
Single muon trigger or di-electron,

eff.0.82
Single muon trigger or
Di-tau trigger, eff.0.38 etau trigger,
eff.0.73 When having a tau-jet L1HLT tau
trigger thresholds are applied 1t jet86GeV,
2t jet59GeV, e-t jet45GeV Offline
selection Leading track pt cut ptgt40GeV, 1 or 3
tracks in cone 0.04 around the leading track, No
other track with ptgt2GeV in the isolation cone 0.4
Efficient against QCD, bb, Wjets
Since b-jets are soft and distributed over the
whole rapidity range, b-tagging efficiency is low
and single b-tagged jet has been requested for
each event, with impact parameter significance
method (3 tracks with sipgt2)
M. Hashemi CERN,CMS (IPM, Iran)
6
Reconstructed Higgs boson mass
em
leptonlepton
leptonjet
jetjet
M. Hashemi CERN,CMS (IPM, Iran)
7
5s discovery contours
With at low luminosity (
) with SUSY parameters values listed, the
5s contour shows that this channel is the most
promising channel for heavy neutral MSSM Higgs
boson discovery.
M. Hashemi CERN,CMS (IPM, Iran)
8
Calculation of statistical and systematic
uncertainties

• The number of signal events that we get after all
  experimental selection cuts is
• For production cross section we have statistical
  and systematic uncertainties
• Systematic uncertainty comes from
• luminosity,
• experimental selection,
• background uncertainties,
• The total uncertainty is the quadratic sum of
  statistical and systematic errors

M. Hashemi CERN,CMS (IPM, Iran)
9
Statistical and luminosity uncertainty

• Statistical uncertainty of the signal events is
  calculated with standard weighted least squares
  procedure by summing over all final states,
  assuming uncorrelated measurements

• The uncertainty on the luminosity measurement is
  assumed to be 5

M. Hashemi CERN,CMS (IPM, Iran)
10
Signal selection uncertainty

• The uncertainty on the signal selection
  efficiency comes from
• The calorimeter energy scale (since jets and
  missing Et thresholds are required)
• Full simulation of di-t jet final states shows
  that we have 2.9 uncertainty on the signal
  selection efficiency assuming 1 uncertainty on
  the calorimeter energy scale.
• b-tagging efficiency
• Using two samples of semileptonic ttbar events,
  single btagging eff. is calculated as the ratio
  of double vs single b-tagged events.
• The statistical uncertainty is 1 which leads to
  2 assumption on the total b-tagging
  uncertainty considering b-jet purity and
  background contribution.
• t-tagging efficiency
• Using two samples of
  and the
  measured ratio of events selected after applying
  tau tagging algorithm and Z mass constraint gives
  2.5 uncertainty on tau tagging efficiency.
• The total selection efficiency error is 4.3.

M. Hashemi CERN,CMS (IPM, Iran)
11
Background uncertainty
The background contribution to the signal
selection efficiency is estimated by fitting the
distribution. The extrapolation of the
background fit to the Higgs mass window gives the
background contribution. By varying the Higgs
mass and tan(b) the background contribution
uncertainty is obtained. Choosing
and tan(b)20 which is close to 5s to
estimate the highest uncertainty, the background
uncertainty is estimated to be The total
systematic uncertainty is 12 comparable with
statistical uncertainty. Now assuming fixed
values for SUSY parameters, the production cross
section can be written as

• Such that the error on tan(beta) measurement is
• Where consists of
• theoretical uncertainties of the production
  cross section and the branching ratio
• cross section uncertainty due to the uncertainty
  on the Higgs boson mass measurement.

M. Hashemi CERN,CMS (IPM, Iran)
12
Theoretical uncertainty on the signal cross
section

• The NLO cross section uncertainty for the signal
  has been shown to be 20-30 for the total
  rate.(Dittmaier, Kramer, Spira, hep-ph/0309204)

22
17

• If a pt cut on both b quarks is applied the
  theoretical error is smaller (10-15), but since
  1-btagging is done the error is taken to be 20.
  
• The branching ratio uncertainty is 3 which is
  due to SM parameters uncertainties.

M. Hashemi CERN,CMS (IPM, Iran)
13
Uncertainty of the mass measurement
Production cross section depends on the Higgs
mass which is measured with some accuracy. This
induces some error on the cross section. The
gaussian fit to the mass distribution is used to
estimate the error in the cross section related
to the Higgs mass reconstruction, The error
induced to the cross section is
estimated by varying the cross section for the
Higgs masses and
At 5s limit when the signal
statistics is lowest, the mass measurement
uncertainty brings 5-6 uncertainty on the tan(b)
measurement.
M. Hashemi CERN,CMS (IPM, Iran)
14
SUSY parameters uncertainty effects
The Higgs sector is sensitive to SUSY parameters
as well as
and tan(b), The SUSY parameters uncertainties are
unknown. To give an estimation of the rate
sensitivity on the SUSY parameters, those were
varied by 20 around the nominal values. The
variation of the rate within the discovery region
is about 11, which leads to at most 6
uncertainty on tan(b) measurement.
M. Hashemi CERN,CMS (IPM, Iran)
15
The statistical uncertainty and the uncertainty
due to the mass measurement is shown in the
table, The total uncertainty includes Statistics,
mass measurement, cross section and branching
ratio, luminosity, experimental selection,
background uncertainties.
M. Hashemi CERN,CMS (IPM, Iran)
16
Small gray errors statistical Large errors
total errors
M. Hashemi CERN,CMS (IPM, Iran)
17
conclusions

• The precision of a tan(b) measurement is
  estimated using
• in CMS with
• The uncertainty includes
• Statistical error
• Error of the Higgs mass measurement
• Theoretical error (cross section and branching
  ratio)
• Luminosity error
• Signal selection uncertainty
• SUSY parameters uncertainties with 20 assumption
  for each
• What is still to be investigated in more detail
• Uncertainty of the SUSY parameters measurement
• Uncertainty of the signal plus background
  selection and background determination
• With signal significance gt5s, the study gives
  better than 35 accuracy on the tan(b)
  determination at the benchmark point considered.
• Errors are dominated by the theoretical errors.

M. Hashemi CERN,CMS (IPM, Iran)