Trileptons final state in SUSY at CMS

Valery Zhukov IEKP Karlsruhe last updated 09.06.2006

Content:

Working group
Talks
Notes
PTDR
Plans
Analysis
Links
Data Samples
Status

New for discussion at CERN 15.03.2006

Working group

University Kalrsruhe, Germany:
Wim deBoer        deboer@physik.uni-karlsruhe.de
Valery .Zhukov zhukov@physik.uni-karlsruhe.de
Martin Niegel       niegel@physik.uni-karlsruhe.de
Anj Vest               vest@physik.uni-karlsruhe.de

Tata Inst. of Fundamental Research,Mumbai,Indoia
Kajari Mazumdar    Kajari.Mazumdar@cern.ch

UC Davis, USA
Maxwell Chertok chertok@physik.ucdavis.edu
Aron Soha
Alexander Stromberg

   Mugla Uni
Kerem Cankocak    Kerem.Cankocak@cern.ch

everybody interested is welkom

Talks

M.Chertok 11.2004
W.deBoer 11.01.2005    cern1101.pdf
V.Zhukov 11.04.2005    fnal1104.ppt
V.Zhukov 12.05.05    cern1205.ppt
K.Mazumdar 06.2005 cern.ps.gz
V.Zhukov     28.07.2005    3lzhukov.ppt
M.Chertok    06.09.2005    lm13l.pdf
V.Zhukov     07.10.2005 tril3l.pdf
V.Zhukov 18.11.2005 susy3lap.pdf
V.Zhukov 16.12.2005    app3l.pdf
V.Zhukov 03.02.2006    anote3l.pdf

Notes

- K.Mazumdar et al. CMS NOTE 97/007
- K.Mazumdar et al. CMSIN 2004/*
- de Boer V. Zhukov M.Niegel    CMS IN 2005/039 tril.ps.gz
- de Boer, Zhukov, Niegel, Sander, Mazumdar   draft CMS AN 2006/003  trilap.ps.gz
- DeBoer, Zhukov et al.    CMS NOTE 2006/113 trilapn.ps.gz

referees comments and answers

PTDR

draft v.1 V.Zhukov: first versiuon july 2005
ptdr_tril1.tex
ptdr_tril1.ps

draft v.1.1: M.Chertok correction august 2005
ptdr_tril2.tex
ptdr_tril2.ps

draft v.2 V.Zhukov: many changes according to the AN 2006/003 february 2006
ptdr.ps
trilept.tex

draft final. V.Zhukov 05.2006
trilept.tar.gz
trilept_scansugra10.pdf
trilept_scansugra50.pdf

Scope of the contribution:

1. Introduction:Trileptons signature and topology(2 and 3 body)
    Direct neutralino-chargino production.
    Mass reconstruction.
    mSUGRA (MSSM) parameter plane and mSUGRA   constrains.
Why trileptons are interesting, comparison with other channels.
2. Signal cross section: Uncertanities (PDF), NLO.
    Cross section background. List of all relevant backgrounds. Uncertanities, NLO.
3.   CMS MC: Full anf fast.
   Reconstruction algorithms specific for the trilepton.
Leptons reconstruction efficiency and purity, momentum and energy(for e) reconstruction.
Jets reconstruction efficiency.
     MET uncertanities from calibrations and corrections.
   Influence of PU and underlying events.
4.   Simulated data smaples.
     Selection cuts and statistics flow.
5. Invariant mass reconstruction.
   Algorithms for Minv construction, combinatorics for the signal and background.
   Dileptons edge reconstruction accuracy. Topology selection.
6. CMS discovery reach.
5 -3 sigma reach in m0-m1/2 for tanb50 and 10 (at least) for mSUGRA
mSUGRA constrains and the discovery reach. Complimentarity to other channels.
Plots, Figures, tables needed:
- 5 sigma discovery reach with complimentary SUSY discovery channels for different tanb.
- Minv mass distributions and illustartion of mass reconstruction for 2 and 3 body cases.
- Event topologies
- Njets for signal and all bkg.ß
- table with the statistics summary

Plans

July 2005	CMS Internal Note (Karlsruhe) tril.ps.gz
July 28 BSM CERN	draft of contribution to PTDR ptdr_tril2.ps
September 6 BSM CERN	review of the plans
October 7 BSM CERN	- Update with L1&HLT trigger study - MET cut - optimization of the leptons, jets and MET cuts - W/Z/gamma+jets and QCD backgrounds(?)
November 18 BSM CERN	- draft of the final CMS note
December 2005	- semi final contribution to PTDR
february 2006	Analysis note AN 2006/003 trilap.ps.gz trilap.tex iCMS, PDTR v2.update ptdr.ps

03.2006

For discussion at CERN 14.03-16.03.2006

1. status in Karlsruhe :
-see last presentation 03.02.2006 and comments
          http://cmsdoc.cern.ch/cms/PRS/susybsm/minutes/feb03-2006.html
-see draft AN 2006/003 note, there is still thing to be added for the fakes and uncertainties
trilap.ps.gz

    MySummary:
     a. Results.
        The main problem is the fakes (mostly electrons) from large c.s. bkg like Zjets and DY.
        Maybe the use of the trimuon state is the only escape. However the investigation
        of the fake leptons rejection is important f.ex. the electron Lhood, lepton isolation, etc.
        can be improved

     b. Selection.
        Neural Network - how it is reliable?
        Main complaint is that the bkg shape after NN is similar to the signal one (as it should be).
I think this a personal question. Nevertheless the stability of the NN selection has to be studied.
        The alternative methodes- Genetic algortihm- has to be explored and compared.
        In the other hand all model dependant selections are a bit dangerous.
I would prefere to have a 'blind' analysis which recons only on the differences from the SM bkg.
        This is especially important for the discover reach scan.
      There is also a room for improvement by using MET(sumET), optimization of lepton Pt cuts, etc.
Since we need a very precise MET reconstruction - a special study is needed,
        striking is a differences between FAMOS and ORCA.

     c. Data samples.
        Situation is not very bright here in respect to the availability.
There are very few published bkg available (ttbar, ZW, WW)
        Some backgrounds are missing, some considered can be improved using more
        advanced data samples (ALPGEN). For the 2->3,4 Z bbar (ZWjets) Kajari is doing it with ALPGEN.
        Most of bkg used in KA are produced on fly with FAMOS and is hard to use with the other analysis framework.

     d. Analysis.
        I used a custom analysis which produces a root tree which can be used further for NN and
        GA selection, i.e. the analysed data samples are not compatible with the ExRoot.
        There are some vice and contra to use this approach.
If we would like to be compatible in the data samples, we have to agree about data format.

     e. Uncertainties and significances.
Time consuming work, especially for reconstructions.
Here all ideas are welkom, CMS is still not setup on this subjects.

2. How to organize efforts, share subjects and results?
     -Manpower and areas for study. In Ka we will continue on the whole subject, more
      concentrating on analysis issues.
   -Timeplanning. The PTDR is going to end soon. The final contribution has to be
carefully finalized, probably using only the existing analysis.
However the work is not over and we need planning for the next 6 months at least.
    In Ka we expect to clearify the fakes problem by this time and produce some
      internal note on this.
     -We also have to define the format of intercommunication and results presentations.
Especially important to solve all conflicted results in between before going to CMS.

4. Preliminary subjects for study:
     -revisit backgrounds. ALPGEN samples
     -NLO contributions and (ALPGEN+MCNLO,COMPHEP PROSPINO) 2->N bkg.
     -ISR/FSR in dilepton bkg.; DY, Zjets, isolation in ORCA(FAMOS), fakes for the ISR(bjets) leptons (Karlsruhe)
     -Trigger. revisit HLT with new PTDR algos.
     -Leptons isolation, identification (electrons mostly) - can be improved?
     -selection criteria, optimization tools.Use of MET, sumET (MET algos)
     -reconstruction uncertainties, detailed consideration for energy scale errors and Pt cuts.
     - ....

10.2005

Needs to be done:
Data samples:
- make a CMKIN 'smart' scan of mSUGRA excl. 3l channel for different tanb
- FAMOS scan of trileptons >10fb region.
- urgent. need ZWjets, QCD jets samples (CMKIN, FAMOS, DST) with ALPGEN(?)
- need a pileup(LL) inclusive large DST sample for LM9, LM7,LM1 for susy bkg.
- need a 3leptons equivalent DST PU (LL) samples for signal
- make all signal and backgrounds publicly available (on demand)
- NLO corrections for the trilepons signature, look more into details
- improve NLO corrections for backgrounds ( MCNLO)

Reconstruction:
- improvement of electron reconstruction efficiency and purity (ORCA and FAMOS), collaborate with Catania.
- electrons bremss recovery by Calo and tracker, (see W.Adams idea).
- optimisation of Pt cuts for muons and electrons; make scan of cuts
- trigger study L1 and HLT (trilepton HLT case). This is valid for ORCA. We need a trilepton HLT menu. FAMOS trigger is still not working.
- optimisation of jets reconstruction efficiency in ORCA and FAMOS), jets cuts and calibrations. Scans of cuts
- combinatorial backgrounds subtraction for signal and backgrounds with muons and electrons efficiencies (OSOF).

Analysis:
- precise 5 sigma scan (FAMOS) , algorithms for signal separation in ivariant mass. Aproximation of bkg and signal. A lot of work here.
- 2-3 body event topology separation: assymetry and corellations, Dalitz plots. A lot.
- mass reconstruction and accuracy; related to first item., fitting accuracies
- complimentarity study. How trileptons are competitve to casade decay (see talk 12.05)

Analysis

07.2005:

The simple example of the FAMOS and ORCA analysis can be found at
(old version)
mysusyfamos062005.tar.gz
mysusyorca062005.tar.gz

It is not exactly the ExRootAnalysis but a simple custom root tree.
The ExRootAnalysis based analysis is also under developments
but probably will make more sense after settling up the DST.
The FAMOS puleup related things can be found at FAMOSPU.html

We collaborate with the Catania group for electron reconstruction .

Links

CMS

BSM SUSY
http://cmsdoc.cern.ch/cms/PRS/results/susybsm/susybsm.html

Other related activity at CMS:
M.Chiorbioli et al semileptonic LM1 dileptons (collaborated), cascade decays.
F.Moortgat et al Heavy higgses (complimentarity), hemisphere separations (important!)
L.Pape et al SUSY region separation (collaborated)
Florida Same Sign leptons

ATLAS
SUSY
http://paige.home.cern.ch/paige/susy/susy.html

FNAL

CDF BSM: http://www-cdf.fnal.gov/physics/exotic/exotic.html

D0 BSM: http://www-d0.fnal.gov/Run2Physics/WWW/results/np.htm

http://www-cdf.fnal.gov/physics/exotic/r2a/20050714.trilepWebPage_hiptmm/
http://www-collider.physics.ucla.edu/cdf/analyses/#

The expected discovery reach for tevatron for different L (green -2fb^-1, blue-10fb^-1,pink-30fb^-1)
Nothing is seen so far (L=0.2fb ^-1)
tevarton

DM search
Karlsruhe Indirect DM search(link to appear)

Data Samples

All DST data samples used in 2005 summer analysis are located locally at Karlsruhe and FZK GridKA (~100 Gb)
The PYTHIA ntuples are availble on demand.
The FAMOS root files (in the specific format) are available on demand.

Status

September 2005
published CMS- IN 2005/039 tril.ps

Juine 2006
CMS AN 2006/003
CMS NOTE 2006/113 cmsnote.pdf

Basic study is finished and terminated.
the follow up questions are under study:
- lepton fakes
- MET at low E scale

contact zhukov@ekp.physik.uni-karlsruhe.de