SUSY Search mSUGRA scenario

1
SUSY Search - mSUGRA scenario
Search for Chargino and Neutralino into 3 leptons
and MET mSUGRA R-parity conserved scenario

• Overview and results for Summer Conferences 2005
• Improvement to the acceptance towards Winter
  Conferences 2006
• Validation of 6.1 data
• Conversion study
• Plan

2
Search overview

• Chargino and Neutralino are a mixed state of the
  SUSY partner of gauge and Higgs boson
• Two production modes
• t channel (suppressed)
• s channel (dominant)

3
Analysis approach
3 isolated and separated leptons large MET
Lepton pt range driven by the mass difference
between chargino and neutralino Large missing
transverse energy due to the escaping neutrino
and massive neutralino (LSP)
First lepton is high pt ? DATASET, high pt single
muon bhmu0d (CMUP18 CMX18) Large Missing
Transverse Energy ? discriminating variable in Rp
searches

• Blind analysis performed as a COUNTING experiment
  
• PURDUE exploring the ???/e channel (to be
  combined to other channels for publication)
• A priori optimization of the event selection
• Blind to data in signal region
• SM prediction tested in 15 control regions
  (invert kinematic cuts)
• Open the box

4
Event selection and results
What are the major backgrounds with signature
3 isolated leptons and large MET ?
Major background Drell Yan and Diboson
(estimated from MC) Fake leptons (estimated
from data)
Blessed Analysis
5
Improvement to the blessed analysis

• Re-investigated dimuon regions with full
  statistics, tuned EWK Z??? and Z??? MC samples
  exploited
• Extended to more channels
• CMIO included as a second muon
• ?e channel (see Elses talk at SUSY Meeting Oct
  18th)
• Good agreement in all 15 control regions
• Acceptance improved (by adding the CMIO channel
  only) by a factor 2 with respect to the blessed
  analysis

6
Combined limit (prelim)
?? channel only
Blessed Analyses (e? included)
Blessed Analyses (e? and CMIO included)
Acceptance up from 0.75 to 1.42
Blessed result 0.37 SUSY events, limit on ??BR
1.2 pb Now 0.70 SUSY events 0 obs ? limit
of 0.62 pb 1 obs ? limit of 0.97 pb
346 pb-1
800 pb-1
7
(No Transcript)
8
Validating 6.1
Validating the first 200pb-1 of 6.1 data
available run range 190697203799 from August
2004 to September 2005

• compare 5.3 data to 6.1 data
• compare 6.1 data to 6.1 MC (on going 6.1
  ntuplizer issues)

• What is different with respect to 5.3 ?
• Compiler switch from KAI to GCC
• Tracking specific improvements
• Include t0 into the track parameters fit
• Lost segments recovery
• Avalaibility of CMX keystone and miniskirt
• How do gen5 and gen6 data compare ?
• Muons and tracks see next slides
• Electron
• CEM E scale lower in 6.1 (0.2 GeV)
• PES 5/9 ratio peaks at higher values in 6.1
• CES dx and dz narrower in 6.1
• Need nvtx correction to the ID
• Jets
• 1 higher scale in gen6 east plug

• For tracks with Ptgt300 MeV
• Determine t0 per primary vertex
• Constraint the tracks to the vertex t0
• Improve chi2/dof, residuals

• Search for hits in missed SLs
• Refit the tracks one extra time
• Reduces the fraction of lost segments in the
  inner superlayers by 90
• Makes the tracking efficiency more uniform vs
  luminosity
• Increases the number of hits/track

9
COT Chi2 and hit distributions

• More COT hits both axial and stereo more layers
• Chi2 peaks at lower values
• Improves due to the event based use of t0
• Tighter cuts might improve fakes rejection

Shown at JPM
10
Dz0 distribution Gen5 vs Gen 6
Mainly due to segments recovery in the inner SL
Shown at Coll Meeting
Post-shutdown 6.1 data
Pre-shutdown 5.3 data
RMS 0.202/-0.004 cm
RMS 0.136/-0.002 cm
11
Momentum Scale
Shown at Coll Meeting
Post-shutdown
Pre-shutdown
Mean 90.80 GeV/c2 RMS 2.497/-0.046 GeV/c2
Mean 90.84 GeV/c2 RMS 2.516/-0.036 GeV/c2
12
CMX Geometry
KEYSTONE
MINISKIRT
MINISKIRT
13
Phi and Eta distributions
TIGHT
CMIO
14
Run dependence

• Current issue understanding the run dependence
  of yield and cross section
• Clear run dependence in W yield and cross section
  less pronounced in Z events
• Similar pattern in electron and muon channels
  luminosity issue?
• Observed run dependence in old data confirmed
  (worse) in new data
• Joint physics task force effort documented at
  JointPhysics/Event Yield Studies

Tight muon selection
Trigger requirement only
15
(No Transcript)
16
Conversion study

• The current algorithm (standard) is based on
  selecting track pairs as conversion candidate
• Need to evaluate the residual conversion (no
  partner track is found)
• Largest reducible background
• Need to improve the rejection power
• Investigating the contribution in details
• ? Compare data and MC in control regions when we
  allow conversion
• ISR not included in the current DY sample
• Need to estimate the contribution
• Compare to the estimate from MC (ISRFSR)
  generated with tight kinematic cuts

17
Tagging efficiency
The relative residual conversion can be estimated
from the conversion efficiency

• MC sample Baur MC Z? with Z??? (ISR FSR)
• Data sample bhmu0d (looked at edil0d
  SUSY_DILEPTON)
• Select Z event to reduce fake contributions with
  large invariant mass window
• Look for loose electron candidate conversion
• (in the future might select conversion data
  sample using the constrain that the inv mass of
  ??e must account for the Z)

18
Kinematics
Partner track might belong to the low pt range
If the loose electron has large Et, the partner
track belongs to the low pt range
---- OBSP level ---- RECO level
Loose Ele from ? conversion
---- OBSP level ---- RECO level
? conversion
Tracking variables used to ID conversion
Dxy
Cotg?
19
Partner tracking efficiency

• Partner tracking efficiency
• ?track probability of finding a partner track
  for a given seed electron
• Calculated in MC
• Select Z??? events with a loose electron
• Identify conversion pair at OBSP level
• Match the loose central electron (seed
  electron) to OBSP electron (?Rlt0.4)
• Look for a reconstructed track matched to the
  OBSP partner of the OBSP electron
• scale factor calculate based on CES information

20
Scale factor (I)

• Need to select a data sample of conversions
• For a given charge of the seed electron, the
  partner track has to be found in the correct
  side in f
• Photon conversion pair electron and positron
  have same Z position at ay radius
• Combined into a criteria to select conversion
  candidates in data

21
Scale factor (II)

• Exploit CES information unbiased and track
  based CES
• Unbiased Ces Cluster seeded from strip/wire
  above threshold
• Track Based Ces Cluster seeded from COT track
• Define the partner track finding efficiency
• on event basis
• as the ratio of the number of events with track
  based conversions over the number of events with
  unbiased conversions
• Calculated in both data and MC take ratio of
  the efficiencies

• Select Z??? events with a loose electron
• For each given seed electron, look for the
  highest CES cluster (unbiased or track based)
  close in Z and in the proper wedge
• ?Z(ZseedEle Zces) lt 4 cm if cluster in proper
  wedge

22
Scale factor (III)

• Preliminary results
• Need to run over the data need higher statistic
  sample
• No background substraction
• Background prompt electrons, fake electrons
• Do not expect large contamination of prompt
  electrons in Z events
• Fake electrons can be estimated (need to be
  consistent when calculate the fake rate itself)
• Any kind of background can be removed using CES
  info the background will populate evenly the
  correct and the wrong wedge close to the seed
  electron

23
Tagging cut efficiency

• Tagging cut efficiency
• ?cut probability of tagging a conversion
• The standard tagging cuts are
• Dxy lt 0.2 cm
• ?cotg? lt 0.04
• Need to select a sample which is independent of
  the above tagging criteria
• Exploit CES information track based CES

• Denominator of the tagging cut efficiency number
  of candidate conversions
• Numerator of the tagging cut efficiency number
  of candidate conversions tagged as conversion
  according to the applied cuts

24
Is the conversion rate reproduced?

• Electron from conversions are removed
• Exploit PYTHIA sample (no ISR)

25
Checks
Electron from conversions are accepted as third
lepton in the event (The fake rate is estimated
WITHOUT removing conversion from the jet
sample) Compare PYTHIA (FSR only) to BAUR
(ISRFSR)
26
ISR-FSR contribution
Compare PYTHIA(FSR) and BAUR(ISR_FSR) with no
generator level cuts and with the same generator
level cuts
BAUR Photon pt gt 5 GeV ?R gt 0.2 M leptons gt
20 GeV M leptons photon gt 20 GeV
27
Summary

• 6.1 data validation (done)
• Run dependence now became task of lum group
  just cross checking
• Additional 440 pb-1 already ntuplized, run over
  it and confirm agreement in the control regions
• When available, will show the validation of the
  Muon Geom in 6.1 at Simulation Meeting
• Conversion (procedure is set)
• Need to generate another sample ?
• More comparison using CdfEm
• Run over high stat data sample (check B group)
• For Winter conferences, use current 5.3 MC and
  6.1 data
• By end of the year,
• Scale factors and trigger eff should be available
  (Top and MultiLepton group)
• (will cross check the results of the trigg
  eff little bit of coding is needed)
• Fake rate (Beate?)
• I will try to look at that
• Jan., run with new scale factors and fake rate
• Febr., sysematics and blessing
• Moriond, second half of March
• Publication with the results shown at Moriond

28
Back up
29
TC results
30
TC control regions - dimuon
31
(No Transcript)
32
TC control regions - dimuon
All dimuons
Mass in G
33
TC control regions - dimuon
All dimuons
2nd muon pT in G
34
Fake rate

• Fake rate calculated without removing conversion
  electrons from the jet sample