TTeruki KamonFor the CMS Collaboration
Mitchell Institute for Fundamental Physics and AstronomyTexas A&M University
&Department of Physics
Kyungpook National University
Ninth Particle Physics Phenomenology Workshop (PPP9)National Central University (NCU), Taiwan, June 3 ~ June 6, 2011
SUSY Searches at CMSJJune 2011 11
Searches at CMS
SummaryOUTLINE3
1) Why? … Dark Matter and SUSY2) Where? … LHC & CMS Detector3) How? … SUSY Searches
PrologueIt has been 13.8 B years, sincethe LHC machine was set up. Themachine finally started providingproton-proton collisions at acenter-of-mass energy of 7 TeVon March 30, 2010 and becamethe energy frontier machine tolead discoveries of new particles.The Standard Model (SM) iscurrently well tested up to ~100GeV, but is expected to breakdown in the TeV domain wherenew physics should occur. This isprecisely the domain that we willstudy at the LHC.
SSUSY Searches at CMSTTeruki Kamon 22
~0.0000001 seconds
Now
CMB
Teruki KKamon
annihilation
combination
LHC
http://www.damtp.cam.ac.uk/user/gr/public/bb_history.html
Dark Matter and SUSY
Probing 10-7 sec. after Big Bang
SUSY Searches at CMS 33
LHC at CERN
27 km ring
Teruki Kamon 44SUSY Searches at CMS
The CMS (21 m x 15 m x 15 m, 12,500 tonnes) is one of two super-fast & super-sensitive detectors,consisting of 15 heavy elements, collecting debris from the collision and converting a visualimage for us. “Particle” Telescope at CERN vs. Hubble Space Telescope in outer space
Compact Muon Solenoid & PF
TTeruki Kamon 55SSUSY Searches at CMS
(see Appendix A)
As of May 31
�Fully hadronic searches:– SUS-10-003 (arXiv:1101.1628, PLB698 (2011) 196) & SUS-11-001: ��T
– SUS-10-011: MET + b-jets + �T
– SUS-10-005: inclusive MET + 3jets– SUS-10-009: Inclusive search with “Razor” variables
�Searches with leptons: – SUS-10-006: MET + jets + single lepton– SUS-10-004 (arXiv:1104.3169): MET + jets + LS dilepton– SUS-10-007 (arXiv:1103.1348, JHEP): MET + jets + OS dilepton.– SUS-10-008: Multileptons– SUS-10-010 & SUS-11-012(~200 pb-1): MET + jets + Z
�Searches with photons: – SUS-10-002 (arXiv:1103.0953, PRL): MET + jets + photons– SUS-11-002 (arXiv:1105.3152): MET + photon + lepton
TTeruki Kamon 66SSUSY Searches at CMS
CMS SUSY Searches in 2010UURL: https://twiki.cern.ch/twiki/bin/view/CMSPublic/PhysicsResultsSUS
EExxaammppllee:: SSUUSSYYg~g~ , q~g~ , or q~q~ production will be dominant, followed
by their decays (e.g., 02��~qq~� ). �� JJeettss
R parity conservation � Stable lightest supersymmetric particle (LSP)
� If LSP is the lightest neutralino ( 01��~ ),
� it will escape the detector �� MMEETT (( TE�� ))� 0
1~�� = Cold Dark Matter candidate �� CCoossmmoollooggyy
� Thus, the evidence of SUSY-like new physics will appearin the Jets+MET final states.
CCoossmmoollooggyy ��� LLHHCC= [Exciting Motivation]��[Right Place&Timing]
Missing ET(& Jets) at the LHC
MET - inferring new physics (e.g., Dark Matter)Teruki Kamon 77SUSY Searches at CMS
TTeruki Kamon 88SSUSY Searches at CMS
http://cdsweb.cern.ch/record/1343076/files/SUS-10-005-pas.pdf
�“All hadronic inclusive” analysis with key variables:– HT = scalar sum of Jet pT (selecting large s-hat production)– MHT = negative vector sum of Jet pT
�Baseline Event selection: – HT Trigger– 3 jet with pT > 50 GeV & |�| < 2.5 (central production)
– Veto events with isolated electrons & muons (suppress EWK background)
– �(MHT, Jet1,2,3) > (0.5, 0.5, 0.3) (reduce QCD background)– HT > 300 GeV & MHT > 150 GeV � baseline selection
�Final Event Selection: – High HT (HT > 500 GeV): High eff. for signals with long cascade
decay chains– High MHT (MHT > 250 GeV): High background rejection
Teruki Kamon 99SUSY Searches at CMS
Analysis Strategy
Baseline selectionw/o MHT cut
Baseline selection
HHT > 300 GeV & MMHT > 150 GeV An out-of-box comparison of Data vs MC for HT and MHT
Baseline Selection
High HTHigh MHT
>150 GeV
>300 GeVMajor BGs:� Invisible Z(�) +
Jets .. Irreduciblebackground
� Top / W + Jets� QCD Jets
Data-driven BG Estimate
Teruki Kamon 110SUSY Searches at CMS
0,0,10tan,250,60]1[
0
2/10
�� �
GeV GeV LM
Amm
Invisible Z
� Three different methods using boson+jets were employed to obtainthe data-driven estimates of this background (substitute boson withMHT)
� Cross check of different channels: “photon + jets” provides anaccurate prediction (Appendix B) � Solely used for limit calculation
remove���� �-������������������������� �
�+
remove�
removeW-
��WWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWWW ��
�-_
– Similar to Z+jets at largepT (MHT)
– High stat (no branchingratio)
– Similar event topology– Higher stat than Z�
(W+jets rate is about x 2.5of Z+jets)
– Lower statistics than �and W
– Suffer fromBr(Z���) / Br(Z�)=1/6
Teruki Kamon 111SUSY Searches at CMS
������ MMET /
MHT
TTeruki Kamon 112SSUSY Searches at CMS
Invisible Z
– FRAG : Correct for fragmentation photons based on JETPHOX– PUR : Correct for photons from light mesons using shower shape
2 1
1 2
jetsdatadatadata/MCMC
Reco MC
jets)Z(data NPURIDFRAG
jetsjets)Z(
N ���� �������
����
���
����
Leptons failing the lepton veto contribute tobackground. There can be 3 reasons to loseleptons:
� the lepton is not reconstructed
� not isolated
� out of acceptance
TTeruki Kamon 113SSUSY Searches at CMS
W �l W/Top
MMET /MMHT
[Step 1] Start with a control sample of eventswith exactly one muon and measure theidentification and isolation (in)efficienciesfrom data
[Step 2] Scale the control sample accordingto the measured (in)efficiencies from data
[Step 1] Start with a muon+jets sample
[Step 2] Replace the muon by tau response template derived from MC
[Step 3] Recalculate HT and MHT including this expected energy from tau
[Step 4] Correct for
– muon acceptance
– Trigger efficiency, Reco efficiency
– BR(W��� )/BR(W��)* BR(��hadrons)
Teruki Kamon 114SUSY Searches at CMS
W/Top
Jet
MET /MHTIdeal Reality
QCDT /
Significant pT imbalance due to� Physics effects: semi-leptonic decay of heavy
flavor quarks� Detector effects:� Intrinsic jet energy resolution� Dead channels (a significant contribution from ECAL)
Balanced multijet Response functions Smearing
[Step 0] Jet response and resolution functions using � + jets and dijetevents
Teruki Kamon 115SUSY Searches at CMS
[Step 1] Rebalance the data events (jets with pT > 10 GeV) using jet pTresolutions by maximizing a likelihood (LJets), being subject toconstraint MHT = 0 � create the pseudo-particle-level QCD events
[Step 2] Smear rebalanced jets (pT > 10 GeV) with resolution functions
� R+S predicts full event kinematics (jet pT and angular distributions)� Consistent with “Factorization Method” (extrapolate two-variable
correlation to signal region)Teruki Kamon 116SUSY Searches at CMS
QCD: Rebalance + Smearing
Baseline selection w/o MHT cut Baseline selection
MHT HT HT
MHT >150 GeV
No excess of observed events over expected StandardModel prediction. Setting limits.
Teruki Kamon 117SUSY Searches at CMS
Results
TTeruki Kamon 118SSUSY Searches at CMS
� MMHT = 693 GeV & HT = 1132 GeV� MMeff = MHT + HT = 1.83 TeV� NNo b-tagged jet & No isolated lepton � IIncompatible with W or top mass� IInvisible Z???
For Future Excitement
HT
MHT
00,10tan
25060]1[
0
2/1
0
��
�
GeV
GeV LM
Amm
� Gluino masses up to ~700 GeV are excluded. Less sensitive to tan���(see Appendix C)� Sensitivity greater than ATLAS at high m0. High HT search region was effective.� Sensitivity lower than ATLAS at high m1/2. Need to look at 2 jet events� See Appendix D for Simplified Model (currently only �3 jets)
Teruki Kamon 119SUSY Searches at CMS
Within the mSUGRA/cMSSM
10tan �
� 44 parameters and a sign: m0, m1/2, tan�, A0, sign(�)– mm0: common mass for “spin 0” particles at the GUT scale
– mm1/2: common mass for “spin 1/2” particles at the GUT scale
10tan �
� CHALLENGE: All hadronicinclusive search iscomplete at the samepace as other searches.
� ROBUST data-driventechniques for all SUSYsearches in 36 pb-1 in2010
� GOOD agreement withthe SM predictions
� HOT: ~1 fb-1/month. Bigexcitement in 2011 &2012
Teruki Kamon 220SUSY Searches at CMS
Summary
• Cross-section limits 0.5 – 30 pb,excluding m(gluino) < 700 GeV inthe mSUGRA/cMSSM plane.
0,0,10tan,250,60]1[
0
2/10
�� �
GeV GeV LM
Amm
TTeruki Kamon 221SSUSY Searches at CMS
Personal Remarks
Excluded by1) Rare B decay b�� s�2) No CDM candidate3) Muon magnetic moment
abc
CDMS II
Rouzbeh Allahverdi, Bhaskar Dutta, Yudi SantosoarXiv:0912.4329
Teruki Kamon
Remark 1: MET + Jets + Taus
Stau - neutralino co-annihilation scenario (e.g., Arnowitt, Dutta, Gurrola, Kamon, Krislock, Toback, PRL100 (2008) 231802)
SSUSY Searches at CMS 222
Remark 2: Bs �� ���
Teruki Kamon SSUSY Searches at CMS 223
jbWbWjttpp )()( ���� ��
jj �ljjjjWpp ��
�l
Remark 3: MET + Jets + W’sBi-Event Subtraction Technique (hep-ph/1104.2508)
B. Dutta, T. Kamon, N. Kolev, A. Krislock
Teruki Kamon SUSY Searches at CMS
�
24
BEST: “jet” mixingfrom two different events
(TTbar, TTbar), (TTbar,W), (W,W)
BEST in TTbar & SUSY
mjj mbW
mjW
mjjTeruki Kamon SSUSY Searches at CMS 225
SummaryInterconnection between Particle Physics and Cosmology
PPC 2011 at CERN, June 14-18PPC 2012 at ???
TTeruki Kamon 226SSUSY Searches at CMS
Remark 4: PPC
CSI: Supersymmetry
at the LHCCollider Scene Investigation
“Simplified” Grand Summary
LHC – keep going!TTeruki Kamon 227SSUSY Searches at CMS
Appendix A: CMS Detector & PF
228SSUSY Searches at CMS
� In this search, all physics objects (jets, leptons, HT, MHT etc) are reconstructed with the PF algorithm.
� Basic idea:– Reconstruct and identify all different types of particles– Apply corresponding calibrations– The list of “particles” is given to the jet clustering and missing ET
(MET) reconstruction algorithm
Teruki Kamon 229SUSY Searches at CMS
Particle Flow (PF) Algorithm
CCharged hadrons~65% of jet energy
Use the high resolution tracker~1% at 100 GeV
Teruki Kamon SSUSY Searches at CMS
PPhotons~25% of jet energy
Use high resolution / good granularity ECALGranularity: 0.02 (�����)Energy resolution: ~2%/�E
Teruki Kamon SSUSY Searches at CMS
NNeutral hadrons~10% of jet energy
Use HCALGranularity: 0.1 (�����)Energy resolution: ~100%/�E
Teruki Kamon SSUSY Searches at CMS
Jet:Charged hadron (solid)Photon (dashed line)Neutral hadron (dotted line)
Particles clustered in jets
TTeruki Kamon SSUSY Searches at CMS
� PPF algorithm improves the performance of jet and missingET reconstruction significantly.
Calorimeter jet
PF jetJet energyresponse
Calorimeter jet
PF jet
Teruki Kamon 334SUSY Searches at CMS
PF Jet and MET Performance
Jet energyresolution
� Baseline selection:
� Consistent with photons and with simulation.� Z � from photon results best precision: therefore solely used
for limit calculation
� Method using leptonic W samples
� Method using leptonic Z samples
HLTRECOIsoW ���� �� ���
���
��
����
��
��
lW
ZRA
NNZNvetobWW
bkgW
obsW)( 10
6 �
��
���
��
lW
ZR
���
���
��
���
�llZ
ZRA
NNZNZZ
bkgZ
obsZ
� )( 02.095.5 ! �
��
���
��llZ
ZR
trigRECOIsolepton ���� ��
2
2
)1(1
)(
HLTtrig
trigleptopnZ
��
���
��
�
where
,
Teruki Kamon 335SUSY Searches at CMS
Appendix B: Invisible Z
4.121(17
)12(32)(1310
168
2918
!�
�����
��
��
:Sim) Z
:Combined
: eeZ
� Less sensitive to tan��
Teruki Kamon 336SUSY Searches at CMS
Appendix C: Large tan� Case
50tan �10tan �
Low tan� vs. High tan�
� Focus on topology instead of underlying physics model (physics is not understood anyway)
� Provide ability to characterize data in model-independent and more comprehensive ways
– Provide intuitive guidance for investigation– Allow one to factorize the key elements potentially present in a
new signal in order to answer specific questions.
� Set limit on cross section (�.Br)� Any model with same topology (parent particle mass,
decay chain, duaghters mass) can be “easily” compared with experimental results.
Teruki Kamon 337SUSY Searches at CMS
Appendix D: Simplified Model
m(gluino) – m(LSP) m(squark) – m(LSP)
TTeruki Kamon 338SSUSY Searches at CMS
Within Simplified Model