Top Mass in the all hadronic channel

1
Top Mass in the all hadronic channel

• P. Azzi, T. Dorigo - Padova University INFN
• Castro, A. Gresele Bologna University INFN
• I. Lazzizzera, S. Amerio Trento University
  INFN
• J. Konigsberg, G. Lungu, S. Sukhanov Florida
  University

• Run I analysis strategy
• Mass Reconstruction in RunI
• Status with RunII MC and data
• Plans

2
Run I analysis strategy

• The all hadronic channel
• BR is large (46) ?
• Events fully reconstructable ?
• Large QCD background S/B is very small ?
• RunI Analysis had to cope with low statistics and
  a low double b-tagging efficiency
• Specific trigger was developed S/B _at_trigger
  1/500
• Two paths for improving S/B
• A)Tight Kinematic gt1 btag S/B 1/4
• used for mass measurement
• B) Loose Kinematic gt2 btag S/B 1/4

3
Run I Analysis Strategy (contd)
Multijet Trigger Data Sample
Large Jet Multiplicity Large Total Transverse
Energy One B-tag jet present

• Kinematic Cuts
• Centrality
• Aplanarity

Two B-tagged jets
Cross Section s7.63.5-2.7 pb
Mass Reconstruction M(t)186105.7 GeV/c2
4
Run I Event Selection for Mass Reconstruction

• Jet counting definition
• Et(raw)gt15 GeV etalt2
• Main background QCD 2?2 processes with extra
  jets due to gluon radiation
• Kinematic very different from tt
• Use data before tag for the background shape
• Signal contamination negligible
• Cuts (corrected jet energies)
• 6?N(jet)? 8
• ?Et? 200 GeV
• ?Et/?s?0.75 (centrality)
• Aplanarity0.0025 x ?3Et ?0.54

ltM(t)gt
175
SET
The ?Et cut has been lowered wrt to the value
optimized for the cross section selection b/c of
its bias on the background shape
5
Mass Measurement in RunI

• Fully reconstructable final state
• Fit the six jets events to tt
  hypotesis (3
  constraints)
• M(jjb)1 M(jjj)2
• M(jj)1M(W)
• M(jj)2M(W)
• Combinations
• No tag 90
• 1 Tag 30
• 2 Tag 6
• Caveat even in events with two tags we required
  only one of the b to be associated to the parton.
  In 30 of the cases the second tag is a charm
  from W decay.
• In Run I we saw no significant improvement in
  top mass resolution from 1 to 2 tags

One b
Two b
6
Mass Measurement in Run I (contd)
Double tagged events (consistency check)
M(t)186105.7 GeV/c2
7
Run I Systematics
Final weights in the combined mass result 0.65
(ljets), 0.19(dilepton), 0.15 (hadronic)
8
The RunII trigger and datasets

• Dedicated trigger TOP_MULTI_JET (after Run
  143938)
• Requires N(jet)gt4 with Etgt15 and ?Etgt125 GeV
• Before Run 143938 L2 was only ?Etgt130 GeV cut
• Total luminosity available now is ?Lint107 pb-1
  (with good run list and SVX requirement)
• Available datasets (Andrea C.)
• Events before Jan shutdown processed with
  4.9.1hpt1
• Full statistics up to May27th processed with
  4.11.1
• Signal MC use official tt MC of the Top Group
  (inclusive)

9
Work in progress and plans

• Top Hadronic Group meets informally every two
  weeks
• Basic philosophy for RunII analysis
• Full characterization of the multi-jet datasets
  (Njetgt4) in terms of kinematical properties and
  heavy flavor content
• Currently we are validating data after the
  shutdown
• Use of new tools (ME MC such as Alpgen) to have
  an extra handle on the background determination
• In Run I no MC was available to properly describe
  high multiplicity events with heavy flavor
• Optimize analysis selection for a S/B1/1 (or
  better) in the signal region
• Work in progress to evaluate a priori how well we
  can do in different luminosity scenarios

10
First look using Run II MC

• Machinery for mass reconstruction is already in
  place.
• Work in progress to add top specific corrections
• Run II Herwig MC
• Jet corrected to Level7
• Compare the resolution for tt in the case of 0, 1
  and 2 tags
• In RunI width was 24 GeV(22 GeV) with 0tag (1
  tag) with full AA corrections.

11
Comparison Run I vs Run II background shape
Entries1153
RUN II multi-jet data 0 tag after kinematical
selection Top specific corrections missing!
RUN I multi-jet data 0 tag after kinematical
selection Full Run I jet corrections
12
Run II signal and background shape

• After a Run I style kinematical selection
• 0 tag (veto) S/B 1/100 , negligible top
  contribution
• gt 1 tag S/B1/6
• gt2 tags S/B1/1

13
Future Plans

• Work in progress to reoptimize analysis to take
  full advantage of the CDF upgraded detector and
  increased statistics
• Need to fully characterize (again)
• Correlations of kinematical and b-tagging
  requirement
• Issues on mass reconstruction from the 2 tags
  requirement
• Background shape?
• improvements in resolution?
• Other ideas bottom/charm difference? Interplay
  of different b-taggers?
• Plan is to have a preliminary results for this
  Winter