YOU ARE DOWNLOADING DOCUMENT

Please tick the box to continue:

Transcript
Page 1: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

R&D STUDIES FOR CMS HE AT THE SUPER LHC CONDITIONS &

INCLUSIVE SEARCH FOR NEW PHYSICS AT CMS WITH JETS AND MISSING MOMENTUM SIGNATURE

Elif Aslı AlbayrakPh.D. Thesis Defense

October 7th, 2011

Page 2: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

2

Outline

Introduction to LHC and CMS Experiment

Part I: Detector Studies

R&D STUDIES FOR HE UPGRADE AT CMS

Part II: Physics Analysis

INCLUSIVE SEARCH FOR NEW PHYSICS AT CMS WITH JETS AND MISSING MOMENTUM SIGNATURE

Page 3: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

LHC and CMS Experiment

Page 4: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

4

The Large Hadron Collider

The LHC is the largest proton-proton (pp) collider designed to run with 14 TeV center of mass energy and 1034cm-2s-1 peak luminosity.

It also provides heavy ion collisions to study the quark-gluon plasma state of the matter.

There are four experiments at LHC

A Toroidal LHC ApparatuS (ATLAS)

Compact Muon Solenoid (CMS)

The Large Hadron Collider Beauty Experiment (LHC-b)

A Large Ion Collider Experiment (ALICE)

The CMS is one of the general purpose experiments, designed to study the physics of pp collisions at 14 TeV at the LHC.

Page 5: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

5

The Compact Muon Solenoid (CMS)

• The CMS is designed to discover Higgs particle and new physics beyond the Standard Model (SM).

Total weight : 12500 TOverall diameter : 15.0 mOverall length : 21.5 mMagnetic field : 4 Tesla

Page 6: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

6

The CMS Detector

The CMS has four main subsystems dedicated to measure the energy, momentum and position of photons, electrons, muons and all the other products of 14 TeV pp collisions:

The magnet

Its bending power allows to determine charge/mass ratio of the tracked particles.

Length/radius ratio and high magnetic field (3.8 T) provides a good momentum resolution.

The muon system

Reconstructing muons, measuring their momentum with a high accuracy and using them for trigger information.

The tracker

Measure charged particle trajectories with high efficiency and provide precise reconstruction of secondary vertices originating from LHC collisions.

The calorimeters

Consists of electromagnetic (ECAL) and hadronic components (HCAL).

Measure the energy of electrons, photons and jets with a high precision.

High accuracy measurement for the missing transverse energy.

Page 7: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

7

The Hadronic Calorimeter (HCAL)

The HCAL is a compact calorimeter and composed of

Barrel (HB)

Endcap (HE)

Forward (HF)

The barrel and the endcaps are sampling calorimeters.

• surround the ECAL and the tracker system.

• cover the pseudorapidity range up to |η| < 3.0.

The forward calorimeter consists of steel absorbers and quartz fibers embedded in it and extends the coverage up to |η| < 5.0.

Page 8: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

R&D STUDIES FOR HE UPGRADE AT CMS

Page 9: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

9

The Hadronic Endcap (HE) Calorimeter

Consists of two large structures at each end of the hadronic barrel detector.

Each HE consists of 14 η towers with 5°φ segmentation.

covers the pseudorapidity region 1.3 < |η| < 3.0, which contains about 34% of the particles produced in the final state.

in the current design 19 layers of plastic scintillators (3.8 mm) are placed between the 7.8 cm brass absorber plates.

Page 10: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

10

Radiation Problem

If LHC discovers the Higgs boson or new physics we will need higher number of events to study rare events such as MSSM Higgs, Higgs coupling to itself.

Higher number of events higher luminosity runs LHC upgrade.

With LHC luminosity upgrade the accumulated radiation will damage the CMS and the other detectors.

Scintillator tiles used in CMS HE will loose their efficiency and stop providing light collection.

As a solution to radiation damage problem, we proposed p-terphenyl (pTp) deposited quartz plates to replace the scintillator tiles.

Advantage: quartz plates are radiation hard.

Disadvantage: light production for quartz plates, photons from Cherenkov process, creates acutely less photons than a scintillation process.

• To increase the light collection efficiency, R&D studies are performed on the quartz plates.

Page 11: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

11

Light Enhancement

Light collection created with Cherenkov process increases with 1/λ2.

More photon can be collected if we use a wavelength shifter method with UV absorption spectra.

For this purpose different wavelength shifters including p-terphenyl(pTp) , 4% gallium dopped zinc oxide (ZnO:Ga), o-terphenyl (oTp), m-terphenyl (mTp) and p-quarterphenyl (pQp) were tested.

Page 12: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

12

Different wavelength shifter (WLS) materials are deposited on quartz plates and the coated plates are tested for the light collection efficiency.

Evaporation and RF techniques are used to deposit the WLS materials on the quartz plates.

Coated plates are prepared at University of Iowa and Fermilab Thin Film Laboratory.

Selection of Wavelength Shifter

Fermilab Thin Film Laboratory ZnO:Ga sputtering system and guns

Fermilab Thin Film Laboratory WLS evaporation

setup

Plain quartz plate

Page 13: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

13

Selection of Wavelength Shifter

Quartz plates coated with different thickness of wavelength shifters are tested at

Fermilab Meson Test Beam Facility (Nov 07 and Feb 08)

CERN H2 area (August 2007)

2 μm pTp deposited quartz plate0.2 μm ZnO:Ga deposited quartz plate

plain quartz plate

Both pTp and ZnO:Ga enhance the light

collection by at least a factor of 4.

Page 14: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

14

Selection of Wavelength Shifter

The pTp and ZnO:Ga enhance the light production by a factor of 4.

oTp, mTp, and pQp did not perform as well as pTp and ZnO:Ga.

Since ZnO:Ga is more difficult to deposit on the quartz plates and does not provide an advantage compared to pTp deposited quartz plates, we decided to focus on pTp.

Page 15: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

15

Radiation Hardness

Different methods were used to test the radiation hardness of pTp

The radiation hardness tests with proton (Indiana University Cyclotron Facility and CERN beam lines) and neutron (Argonne).

The 90Sr activated light outputs of pTp samples before and after irradiation were compared (University of Mississippi CMS Laboratories).

16% light output lost after 200 kGy of proton irradiation.

After 200 kGy radiation damage level slows down.

After 400 kGy still have more than 80% of the initial light collection

Page 16: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

16

The Calorimeter Capabilities

A calorimeter prototype was built with 2 μm pTp deposited (one side) quartz plates .

The 15cm x 15cm x 5mm quartz plates are used.

For hadronic (electromagnetic) configuration 7cm (2cm) absorbers were used between each layer.

The prototype was tested for

hadronic capabilities with 30, 50, 80, 130, 200, 250, 300, and 350 GeV pion beams.

electromagnetic capabilities with 50, 80, 100, 120, 150 and 175 GeV electron beams.

Page 17: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

17

Hadronic Capabilities

Detector linearity Hadronic Energy ResolutionLongitudinal Shower Profile

DataGeant4 simulations

DataGeant4 simulations

1% hadronic response linearityA good agreement betweendata and simulation.

Solid line -> simulationPoints -> data

•1% hadronic response linearity.1% hadronic response linearity.•15% hadronic resolution at 350 GeV pion beam.15% hadronic resolution at 350 GeV pion beam.•On a bigger scale it can reach up to current HE resolution. On a bigger scale it can reach up to current HE resolution. •8% at 300 GeV pion beam.8% at 300 GeV pion beam.

Page 18: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

18

Electromagnetic Capabilities

Detector linearityElectromagnetic Energy

ResolutionLongitudinal Shower Profile

DataGeant4 simulations

DataGeant4 simulations

3% em response linearityA good agreement betweendata and simulation.

Solid line -> simulationPoints -> data

•3% electromagnetic response linearity.3% electromagnetic response linearity.•Above 120 GeV both simulation and data converge to 5.6%Above 120 GeV both simulation and data converge to 5.6%• It can be used as a radiation hard EM calorimeter for It can be used as a radiation hard EM calorimeter for future colliders.future colliders.

Page 19: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

19

Results

Test beam results and Geant4 simulations showed that pTp deposited quartz plates are perfect candidates to replace the current HE scintillator tiles. Both quartz and pTp radiation hard and cost efficient

pTp deposited material loses only 20% of the initial light collection after 400 kGy proton irradiation.

well above higher luminosity conditions (25Mrad = 250 kGy)

pTp deposited quartz plates increase the light yield by at least factor of 4.

The pTp deposited quartz plate calorimeter is a good option in terms of accomplish the current HE calorimeter performance.

Page 20: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

20

INCLUSIVE SEARCH FOR NEW PHYSICS AT CMS WITH JETS AND MISSING MOMENTUM SIGNATURE

Motivation

Supersymmetry

Analysis

Data Driven Background Estimations

Result and Interpretation

Conclusion

Page 21: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

21

Motivation

The Standard Model (SM) can explain our nature’s working mechanism with a high accuracy but there are still unanswered questions such as

Why some force carriers have mass but others do not?

How does the electroweak symmetry breaking mechanism work?

Can gauge couplings be unified at a high mass scale?

What is the source of dark matter in the universe?

Many beyond SM physics theories such as Supersymmetry, extra dimensions, Technicolor, and fourth family try to address these questions.

Supersymmetry (SUSY) is favorite explanation for most of the theorist because it can lead to incorporation of gravity to particle physics

Page 22: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

22

The Supersymmetry (SUSY)

Sfermions Spin 1Sfermions Spin 1 Sbosons Spin Sbosons Spin 1/21/2

Squarks

Sle

pto

ns

SU

SY Fo

rce C

arrie

rs

SUSY ParticlesSUSY Particles

Fermions Spin 1/2Fermions Spin 1/2 Bosons Spin Bosons Spin 0,10,1

Qu

arks

Lepto

ns

Force

Carrie

rs

Standard Model ParticlesStandard Model Particles

SUSY is a symmetry that relates fermions and bosons.

Introduces a spectrum of new particles which are the superpartners of SM particles.

Superpartners have the same masses (unbroken symmetry) and quantum numbers with SM particles but differ by half spin difference.

Sparticles are not observed in nature ➜ SUSY must be broken.

Page 23: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

23

The Minimal Supersymmetric SM (MSSM)

Minimal extension of the SM with minimal particle content.

Respects the same SU(3)C x SU(2)L x U(1)Y gauge symmetries as does the SM.

Assumes that the interaction between particles conserves R-parity.

R = (-1)3(B-L)+2S, which is a multiplicative quantum number with spin S, baryon number B, and lepton number L

All the superpartners are created in pairs.

The lightest supersymetric particle (LSP) is stable and weakly interacts with particle.

LSP is a candidate for the cold dark matter in the universe.

Page 24: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

24

Experimental Signature

Multijet events with large missing momentum is the most generic experimental signature for R-parity conserving SUSY. Long cascade decays of

sparticles ⇒ multijets

LSP will escape the detector ⇒ missing momentum

Page 25: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

25

INCLUSIVE SEARCH FOR NEW PHYSICS AT CMS WITH JETS AND MISSING MOMENTUM SIGNATURE

Motivation

Supersymmetry

Analysis

Data Driven Background Estimations

Result and Interpretation

Conclusion

Page 26: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

26

Inclusive Search for New Physics

Multijet events with a large missing transverse momentum (MHT) search for 36 pb−1 of pp collision data collected with the CMS detector from 2010 March to 2010 November.

The event selection starts with a loose requirement (baseline selection). Later on tighter requirements are applied in order to define search selections.

The variables used in this analysis

MHT: magnitude of the negative vectorial sum of the transverse momenta of the jets with pT > 30 GeV and |η| < 5.0

HT: scalar sum of the transverse momenta of the jets with pT > 50 GeV and |η| < 2.5.

Page 27: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

27

The SM Backgrounds

The largest SM backgrounds for multijet and large MHT analysis are coming from Z(→ )+jets

W+jets and ttbar

QCD multijet events with large missing momentum from leptonic decay of jets

There are also contributions due to jet mismeasurements, and noise or dead components from the detector

Page 28: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

28

Simulations

Monte Carlo (MC) simulated background and signal samples produced with a detailed Geant4 CMS detector simulations are used in this analysis.

The predicted cross sections for MC samples are normalized to next-to-leading (NLO) or next-to-next-to-leading-order (NNLO) cross sections when available and the event yield is normalized to total integrated luminosity of 36 pb−1.

Background Samples

The QCD multijet, ttbar, W, Z, γ+jets, dibosons and single top samples are generated with PYTHIA and MADGRAPH generators.

Signal Sample

The LM1 (CMS low mass SUSY point), with constrained MSSM (CMSSM) parameters m0 = 60 GeV, m1/2= 250 GeV, A0 =0, tanβ =10, and sign(μ) > 0, is used as our benchmark point.

Page 29: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

29

Event Selection

Data used in this analysis are collected with HT trigger, defined as the scalar sum of the transverse momenta of the calorimeter jets with pT > 20 GeV.

At least 3 jets with pT > 50 GeV and |η| < 2.5.

|∆φ(jet1,2,[MHT)| > 0.5 rad and |∆φ(jet3,[MHT)| > 0.3 rad (to reduce QCD background).

Veto on isolated muons and electrons (to reduce EWK background).

HT > 300 GeV and MHT > 150 GeV (baseline selections)

Search Selections

MHT > 250 GeV (High MHT), motivated by the search for generic dark matter candidate coupled with high background rejection.

HT > 500 GeV (High HT), sensitive to higher object multiplicities like SUSY cascade decays.

Page 30: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

30

MHT and HT in Data and MC

MHT HT

BeforeMHT > 150

GeVcut

AfterMHT > 150

GeVcut

Page 31: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

31

Event Yield in Data and MC

Event yield after trigger selection, cleaning and event selections are

The simulations are only used for verification.

All the SM backgrounds in this analysis are estimated by using data-driven methods.

Page 32: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

32

INCLUSIVE SEARCH FOR NEW PHYSICS AT CMS WITH JETS AND MISSING MOMENTUM SIGNATURE

Motivation

Supersymmetry

Analysis

Data Driven Background Estimations

Result and Interpretation

Conclusion

Page 33: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

33

Data Driven Background Estimations

Z( ) + jets

W and ttbar

QCD

Rebalance and Smear Method (R&S)

Jet Resolution Measurements

γ + jet pT balance

• Gaussian Measurements

• Method

• Uncertainty Measurements

• non-Gaussian Measurements

Combination with dijets

Results of R&S Method

Page 34: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

34

Z( )+jets Background Estimation

Irreducible background to jet + MHT search.

Estimated by using γ+jet sample at high pT.

Similar electroweak correspondence between Z boson and γ.

σ(Z→ +jets)/σ(γ+jet) provides a good handle to estimate MHT spectrum.

Only direct photons are related to Z production.

Contribution from fragmentation photons and isolated neutral pions and η mesons are background

Number of predicted events for Z→ +jets are calculated by multiplying the number of γ+jets events with Z/γ correction factors.

Page 35: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

35

W and ttbar Background Estimation

The W+jets and ttbar events are not rejected by the lepton veto if

a lepton from W or top-quark decay is

outside the geometric or kinematic acceptance

not reconstructed

not isolated

a tau lepton decays hadronically (τh).

The sum of lost-lepton and τh W+jets and ttbar.

lost lepton

Page 36: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

36

The W/ttbar ➞ e,μ+X (Lost Lepton)

Events where a W boson decays leptonically and not rejected by the lepton veto.

To estimate the number of events in the signal region a control sample with exactly one well identified and well isolated muon is used and corrected for non-isolated, non-identified and not-accepted leptons.

control region:isolated, identified

and accepted leptons

signal region:non-identified leptons

signal region:non-isolated

leptons

signal region:not-accepted leptons

Page 37: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

37

The W/ttbar ➞ τh + X

A control sample with exactly one muon with pT > 20 GeV and |η| < 2.1 is used to estimate the hadronic tau background.

Each muon in the control sample is replaced by a tau jet.

The control sample is corrected for kinematic and geometric acceptance of the muons

muon trigger, reconstruction and isolation efficiencies

relative branching fractions of W decays into muons or hadronic tau jets.

Page 38: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

38

QCD Background Estimation

Due to its high production cross-section QCD multijet events are the largest background to jets + MHT final state.

A well balanced QCD multijet event at parton level can be reconstructed with a significant missing energy due to

large fluctuations in the calorimeter response to the energy of jets

the semi-leptonic decays of heavy flavor quarks

dead or malfunction channels

These effects manifest themselves as deviation from Gaussian nature of jet resolutions.

It is very important to measure the complete response functions of jets (Gaussian core + non-Gaussian tails) to be able to estimate the QCD background to jets+MHT final state from the data.

Page 39: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

39

Rebalance & Smear Method

missing momentum

Jet1

Jet2

Jet3

Rebalance:Response function

(Gaussian core)

Seed Events:A good estimation for the true

well balanced QCD events (MHT ~ 0).

Jet2

Jet1

Jet3

missing

momentum

Smear:Response function

(Full response)

Inclusive QCD Prediction:Smeared seed events with high missing momentum.

Response functions are measured by two different approaches based on pT balance.

lower pT region: γ+jet pT balance method

higher pT region: dijet asymmetry method

Measured resolutions are used to derive Data/MC correction factors depending on jet pT and jet η.

Jet1

Jet3

Jet2

missing

momentum

Reconstructed Events:QCD+ttbar+w+Z+...+BSM

Page 40: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

40

γ+Jet pT Balance Technique

CMS reconstruct photons with an excellent energy resolution (∼ 1%).

σ(pTJet/pT

γ) a good estimator of the jet pT resolution.

intrinsic(MC truth) imbalance

Varies as a function of SecondJetPt

independent from the extra jet activity in the

event.

Page 41: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

41

Dataset and Event Selections

γ+jet resolution measurements are based on 36.1 pb−1 of pp collision data collected during 2010.

γ+jet data is collected with different trigger paths based on γ pT.

Events with well identified photons are used in this analysis.

HCAL, ECAL, tracker isolations, and shower shape requirements are applied to the data to suppress QCD background.

Events with a track seed in the pixel detector are vetoed to discriminate photons from electrons

leading jetγ

pT > 10 GeV, |η| < 1.3

|Δϕ(γ,Jet)|> 2.7 resolutions are measured in resolutions are measured in different different ηη and p and pTT bins depending bins depending

on on γγ p pTT and leading jet and leading jet η η in the in the events.events.

Page 42: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

42

Measurement of Jet Resolutions: γ+Jet

To study additional jet activity resolutions are measured in the fraction of pT

Jet2/pTγ.

Within the statistics MC well

predicts data.

Reducing secondary jet

activity narrows measured

distributions.

Intrinsic resolution is independent of

any additional activity in an

event.

component of jet resolution as a function of pT

Jet2 / pTγ

Resolutions are measured both in data and MC for

back-to-back γ+jet events .

Page 43: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

43

Method to Derive Data/MC Ratio

Intrinsic resolutions are calculated both for data and MC.

Assumption : imbalance component in data is same as MC and subtract this same component from measured resolutions in quadrature for various bins of pT

Jet2 / pTγ .

measured resolutions intrinsic resolutions Data/MC ratio

intrinsic resolution is expected to be flat in MC and therefore is fitted with a zero degree polynomial before Data/MC ratio is

calculated.

Page 44: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

44

Data/MC Ratio for Intrinsic Resolutions

The measured ratio is consistent with being independent of pT and is fitted with a zeroth degree polynomial.

•To complete resolution measurements for Gaussian core, the following systematics uncertainties on the measured Data/MC ratio were studied; Variation of the extrapolation fit range.• Effect of |Δϕ(γ,Jet)| requirement• Uncertainty of jet energy corrections.• Particle level imbalance • Flavor difference between γ+jet and dijet events • Pileup subtraction.

For each item the Data/MC ratio is recalculated and the relative difference

between the new and the nominal ratio is used in systematical uncertainty

assignments.

Page 45: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

45

Variation of Extrapolation Fit Range

Data/MC ratio for intrinsic resolutions is recalculated for the two fit regions and relative difference between two ratio are used to define uncertainty.

0.0 <|η|< 1.1

1.1 <|η|< 1.7

1.7 <|η|< 2.3

2.3 <|η|< 3.5

Higher ExtrapolationRange 1.079 1.108 1.076 1.184

ΔR (uncertainty up) 0.009 0.003 0.01 0.002

Lower ExtrapolationRange 1.065 1.095 1.061 1.169

ΔR (uncertainty down) 0.005 0.007 0.005 0.0129

Page 46: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

46

Effect of |Δϕ(γ,Jet)| Requirement

Resolutions are measured for back-to-back γ+jet events with |Δϕ(γ,Jet)|> 2.7 rad.

To study the effect of this requirement resolutions are measured in MC for two different Δϕ selections (2.1 and 2.7 rad) and their ratio is calculated.

Since the deviation from 1 on the measured MC(1)/MC(2) ratio is less than 1% the effect of ∆φ requirement is not

taken into account in systematical error calculations.

Page 47: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

47

Uncertainty of Jet Energy Corrections

To determine the systematical uncertainty due to JEC, the jet energy corrections applied on the reconstructed jets are varied by up and down with the official uncertainties provided by CMS JEC group.

0.0 <|η|< 1.1

1.1 <|η|< 1.7

1.7 <|η|< 2.3

2.3 <|η|< 3.5

JEC up 1.080 1.106 1.104 1.189

ΔR (uncertainty up) 0.01 0.004 0.038 0.006

JEC down 1.044 1.078 1.042 1.123

ΔR (uncertainty down) 0.026 0.024 0.024 0.06

Page 48: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

48

Particle Level Imbalance

MC imbalance resolution is used to calculate intrinsic resolutions in data (➩ assuming MC closely describes data).

Various sources can lead to different imbalance component in data and MC

1. treatment of multi-parton final states: Different MC event generators (PYTHIA, HERWIG, MADGRAPH) are used.

2. modeling of hadronization: The hadronization parameter in PYTHIA is turned on and off.

3. modeling of kT kick: 1 GeV smearing of second particle jet pT is studied.

0.0 <|η|< 1.1

1.1 <|η|< 1.7

1.7 <|η|< 2.3

2.3 <|η|< 3.5

Multi-parton Final State 0.017 0.03 0.064 0.001

Hadronization 0.009 0.005 0.023 0.038

Modeling of kT Kick 0.013 0.014 0.021 0.024

Assigned Uncertainty ±2% ±3% ±4% ±4%

Relative changes on the measured Data/MC ratio

Page 49: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

49

Flavor Difference Between γ+jet and Dijet Events

Intrinsic resolutions are measured in γ+jet and QCD MC samples and their ratio is calculated.

For the 20 - 300 GeV pT region where the resolutions are measured from γ+jet

sample, the intrinsic resolutions differ by ∼ 3%.

If MC is wrong by 30% (conservative assumption) this ratio can vary from 2-4%

Since the final measurement is Data/MC ratio, common biases and systematics

cancel and residual uncertainty becomes ~1%.

This systematic is included on measured Data/MC ratio when γ+jet and dijet results

are compared.Photo

nJe

t /

QC

D

Page 50: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

50

Pileup Subtraction

To study the effect of additional large number of soft pp collisions on measured Data/MC ratio

Data/MC ratio is recalculated with a data sample where jets are additionally corrected for pileup.

Since the MC samples were produce without pileup effect the jets in MC sample are not corrected for pileup.

without pileup correction with pileup correctionAssigned

Uncertainty

0.0 <|η|< 1.1 ±4%

1.1 <|η|< 1.7 ±2%

1.7 <|η|< 2.3 ±2%

2.3 <|η|< 3.5 ±5%

Since the data sets used in this analysis are not corrected for additional pileup, the assigned uncertainties are not included in the final Data/MC ratio.

Page 51: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

51

Measurement of Jet Resolutions: γ+Jet

Contributions from extrapolation, JEC and particle level imbalance are accepted to be uncorrelated and final systematical uncertainty is calculated as the quadrature sum of the individual components.

Page 52: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

52

Estimation of non-Gaussian Component

The resolutions in the MC sample are stretched according to the measured Data/MC ratio.

Resolutions measured in data and MC are compared, and the number of events outside the 2.5 σ range is counted in the bins of γ pT to compare resolution tails.

Not enough statistics to measure the resolution tails with precision

A constant fit to the ratio is consistent within the errors with the study using dijet sample, and provides a cross check.

Page 53: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

53

Combined Measured Resolutions

Dijet asymmetry is measured as (pTJet1 - pT

Jet2)/(pTJet1 + pT

Jet2) as a function of jet η and jet pT.

An extrapolation to no second jet activity is performed to eliminate secondary jet activities.

Data/MC ratio measured in γ+jet and dijet samples.

Combined Data/MC ratio in various η ranges.

Data/MC correction factor for resolution tails.

Page 54: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

54

Results of R&S Method

The measured Data/MC ratios from data driven jet resolution measurements are used to correct the MC truth resolution.

Corrected MC truth resolutions are used to smear the jets in the seed sample.

Smeared seed events are used

to estimate kinematic distributions (MHT and HT) of QCD sample

to predict the number QCD background after event selections

R&S method on MC

R&S method on data

Page 55: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

55

INCLUSIVE SEARCH FOR NEW PHYSICS AT CMS WITH JETS AND MISSING MOMENTUM SIGNATURE

Motivation

Supersymmetry

Analysis

Data Driven Background Estimations

Result and Interpretation

Conclusion

Page 56: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

56

Results

No excess is observed over predicted number of events ➡ Limit

Settings

Page 57: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

57

Within the CMSSM

tanβ=10, μ>0, A0 =0

Depending on squark and gluino mass a 95% CL upper limit on the production Depending on squark and gluino mass a 95% CL upper limit on the production cross section is obtained for 2-3 pb range. cross section is obtained for 2-3 pb range.

Gluino masses below 500 GeV are excluded at 95% CL for squarks with mass Gluino masses below 500 GeV are excluded at 95% CL for squarks with mass below 1 TeV.below 1 TeV.

Page 58: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

58

Within the Simplified Model Spectra

Simplified Models remove the complexity of the physics models with several parameters.

Characterize the experimental data in terms of a small number of basic parameters.

With simplified models, the experimental results can be translated to any desired framework.

Two benchmark simplified models are studied for jets+MHT signature.

pair produced gluinos, where each gluino decays to two light quarks and the LSP

pair produced squarks, where each squark decays to one jet and the LSP.

The simplified model samples are simulated with PYTHIA generator for a range of particle masses which are involved in the decays of chosen benchmarks.

Page 59: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

59

Within the Simplified Model Spectra

High MHT Selection

A 95% CL exclusion limits on new particle production cross A 95% CL exclusion limits on new particle production cross section for 0.5-30 pb range, depending on the masses of new section for 0.5-30 pb range, depending on the masses of new

particles.particles.

Page 60: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

60

INCLUSIVE SEARCH FOR NEW PHYSICS AT CMS WITH JETS AND MISSING MOMENTUM SIGNATURE

Motivation

Supersymmetry

Analysis

Data Driven Background Estimations

Result and Interpretation

Conclusion

Page 61: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

61

Conclusion

A search on new physics with jets+MHT signature based on 2010 pp data collected with CMS at LHC is performed.

Data driven methods are used to measure jet energy resolutions and estimate SM backgrounds.

Data showed a good agreement with SM predictions.

No excess is observed over predicted background.

Upper limits are derived in the R-parity conserving CMSSM with A0 = 0, μ > 0, and tanβ=10.

Gluino masses below 500 GeV are excluded for squarks with mass less than 1 TeV (CMSSM).

Upper limit on 2-3 pb range production cross section.

Set limits on σxBr for simplified models.

Upper limit on 0.5-30 pb production cross section depending on the new particles’ mass.

Page 62: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

Backup Slides

Page 63: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

63

Simplified Models: Selection Efficiency

The efficiencies are shown as a function of gluino/squark and LSP mass.

The models are only valid when the gluino mass is greater than the LSP mass.

The signal acceptance increases with higher mass splitting.

The acceptance is low on diagonal

mass splitting is low

jets are produced with low transverse momentum.

MHT > 250 GeV, HT > 300 GeV

MHT > 250 GeV, HT > 300 GeV

MHT > 150 GeV, HT > 500 GeV

MHT > 150 GeV, HT > 500 GeV

Page 64: Elif Aslı Albayrak Ph.D. Thesis Defense October 7 th , 2011

64

Simplified Models: Experimental & Theoretical Uncertainties

experimental experimental experimental experimental

theoretical theoretical theoretical theoretical

The sources of the systematic uncertainties Experimental Uncertainties: jet energy scale and resolution, the lepton veto, the

cleaning, and the trigger selection.

Theoretical Uncertainties: The initial and final state radiation, and the parton distribution functions


Related Documents