Polaroid jetography an album of jet physics measurements and - - PowerPoint PPT Presentation

▶
polaroid jetography
SMART_READER_LITE
LIVE PREVIEW

Polaroid jetography an album of jet physics measurements and - - PowerPoint PPT Presentation

Oct 18, 2023 418 likes •1.06k views

Polaroid jetography an album of jet physics measurements and searches at the ATLAS experiment Caterina Doglioni 1 1 University of Geneva HEP Seminar, University of Virginia - 17/09/13 Introduction Why jets? Large Hadron Collider: quark and

slide-1
SLIDE 1

Polaroid jetography

an album of jet physics measurements and searches at the ATLAS experiment Caterina Doglioni 1

1University of Geneva

HEP Seminar, University of Virginia - 17/09/13

slide-2
SLIDE 2

Introduction

Why jets? Large Hadron Collider: quark and gluon (→ jet) factory

1

Use jets for measurements:

understand QCD (backgrounds), test reconstruction and calibration performance

2

Use jets for searches:

probes for new physics

Why jetography? Main message of the day: there’s many ways to make a jet

(see G. Salam’s primer)

Why polaroid? I only have limited time... This talk: quick snapshots of large ATLAS jet physics program

2 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-3
SLIDE 3

Outline

Overview of jet reconstruction: jet finding, calibration, performance Selected ATLAS results on jet physics: measurements and exotics searches

1 Overview of the ATLAS detector 2 Introduction to jets

Introduction to jet algorithms Jet Algorithms in ATLAS

3 Jet substructure

Introduction Jet substructure performance

4 Jet performance

Jet calibration Jet energy scale uncertainty Jet resolution

5 Standard Model jet results

Jet triggers Measurement of jet properties Jets, dijets and multijets

6 Searches with jets

Dijet analysis Photon+jet analysis Mono-X analyses for dark matter

3 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-4
SLIDE 4

The ATLAS detector

4 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-5
SLIDE 5

Overview of the ATLAS detector

The ATLAS Detector in 2012

Excellent performance of the LHC and of the ATLAS experiment: 5 and 21 fb−1 of pp data recorded in the 7 and 8 TeV runs

+ heavy ion / p − P b data (not covered here)

263 papers published, 530 public notes and counting

Month in 2010 Month in 2011 Month in 2012 Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Peak interactions per crossing 5 10 15 20 25 30 35 40 45 50

= 7 TeV s = 7 TeV s = 8 TeV s

ATLAS Online Luminosity

2012 challenge: high luminosity

Multiple interactions per bunch crossing → optimize trigger, object reconstruction

5 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-6
SLIDE 6

Overview of the ATLAS detector

The ATLAS Detector

For the measurements described in this talk: inner detector, calorimeter system

6 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-7
SLIDE 7

Overview of the ATLAS detector

The ATLAS inner detector and calorimeters

Inner detector

Pixel detectors, semiconductor tracker (SCT), transition radiation tracker

≈ 87M readout channels, coverage up to |η| <2.5 Immersed in 2T magnetic field from solenoid

Electromagnetic and hadronic calorimeters

Subsystem technology and granularity ↔ shower characteristics

transverse and longitudinal sampling very fine granularity: ≈ 200 000 readout cells up to |η| <4.9

Energy deposits grouped in noise-suppressed 3D topological clusters

noise definition includes pile-up and electronic noise

7 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-8
SLIDE 8

Jet algorithms: basics

8 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-9
SLIDE 9

Introduction to jets – Introduction to jet algorithms

Chaos from order, order from chaos?

A high-pT dijet event: how we see it

...from the back of an envelope...

9 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-10
SLIDE 10

Introduction to jets – Introduction to jet algorithms

Chaos from order, order from chaos?

A high-pT dijet event: how we see it

...according to QCD from a MC generator...

I cheated: this is a semileptonic t¯ t event from MCViz, but you get the idea

9 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-11
SLIDE 11

Introduction to jets – Introduction to jet algorithms

Chaos from order, order from chaos?

A high-pT dijet event: how we see it

...in the ATLAS calorimeter...

Note: some ’cleaning’ already performed: ATLAS topological clustering algorithm

9 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-12
SLIDE 12

Introduction to jets – Introduction to jet algorithms

Chaos from order, order from chaos?

A high-pT dijet event: how we see it

...after applying a jet algorithm. Need algorithms to define jets out of underlying constituents

9 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-13
SLIDE 13

Introduction to jets – Introduction to jet algorithms

Jet algorithms: basics

Goal: kinematics of jet ↔ kinematics of underlying physics objects Use a jet algorithm to cluster objects into a jet

From M. Cacciari, MPI@LHC08

Apply same jet definition to objects on different levels:

1

Partons

2

Particles →Truth Jets

(only particles from the hard scattering)

3

Calorimeter objects (ATLAS: Towers, Topoclusters) → Reconstructed Jets

4

Tracks → Track Jets

10 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-14
SLIDE 14

Introduction to jets – Introduction to jet algorithms

Jet algorithms: basics

Goal: kinematics of jet ↔ kinematics of underlying physics objects Use a jet algorithm to cluster objects into a jet

From G. Salam, MCNet School 2008

Apply same jet definition to objects on different levels:

1

Partons

2

Particles →Truth Jets

(only particles from the hard scattering)

3

Calorimeter objects (ATLAS: Towers, Topoclusters) → Reconstructed Jets

4

Tracks → Track Jets

10 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-15
SLIDE 15

Introduction to jets – Introduction to jet algorithms

Wishlist for jet finding algorithms

No right jet algorithm Different processes ↔ different algorithms / parameters (we’ll see more of this later...) Requirements:

  • 1. Theoretically well behaved → no αs dependence of jet configuration:

Infrared safety Collinear safety

  • 2. Computationally feasible → fast
  • 3. Detector independent

11 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-16
SLIDE 16

Introduction to jets – Introduction to jet algorithms

More safety warnings

Crucial to analyse data with infrared / collinear safe jet algorithm! Theory matters:

From G. Salam, MCNet School 08 12 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-17
SLIDE 17

Introduction to jets – Introduction to jet algorithms

Implementation of jet algorithms

Goal: kinematics of jet ↔ kinematics of underlying physics objects Use a jet algorithm to cluster objects into a jet Basic algorithm: event display + physicist “Everyone knows a jet when they see it” Note: don’t try this at home when the LHC is running ...but what is really needed for communicating results:

1

full specification of algorithm and parameters → how to group objects

2

recombination scheme → how to merge objects characteristics

3

treatment of overlapping jets (if any) → how to avoid double counting

13 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-18
SLIDE 18

Introduction to jets – Jet Algorithms in ATLAS

Jet algorithms available in ATLAS

Cone-based algorithms

Cone in y − φ space around object momentum vector Jet = objects in cone Available on the (ATLAS) market:

ATLAS Cone unsafe! Seedless Infrared Safe Cone (SISCone)

Sequential recombination algorithms

Group objects based on minimum relative distance Jet = grouped objects Available on the (ATLAS) market:

Kt Cambridge-Aachen Anti-Kt

What algorithms for data?

From G. Salam, MCNet School 2008 14 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-19
SLIDE 19

Introduction to jets – Jet Algorithms in ATLAS

Sequential recombination algorithms (kt-like)

Algorithm specification: kt

di,j = min(p2

T,i, p2 T,i) ∆R2

D2 ; di,Beam = p2

T,i

D : algorithm parameter (≈ weight for angular distance ∆R) Iterate:

1

For every pair of objects i, j calculate dmin = min(di,j, di,beam)

2

If dmin = di,j recombine objects Else i is a jet, remove it from list a Recombination starts from soft objects

aATLAS default: inclusive algorithm

Idea:

y p

T

A B C min A,B

a)

y p

T

AB C

b)

y p

T

AB C

c)

jet jet d = d

min C,Beam

d = d

min AB, Beam

d = d y p

T

AB C

d)

15 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-20
SLIDE 20

Introduction to jets – Jet Algorithms in ATLAS

Sequential recombination algorithms (kt-like)

Algorithm specification: Cambridge- Aachen

di,j = ∆R2 D2 ; di,Beam = 1 D : algorithm parameter Iterate:

1

For every pair of objects i, j calculate dmin = min(di,j, di,beam)

2

If dmin = di,j recombine objects Else i is a jet, remove it from list a Distance-based recombination

aATLAS default: inclusive algorithm

Idea:

y p

T

A B C min A,B

a)

y p

T

AB C

b)

y p

T

AB C

c)

jet jet d = d

min

d = 1 y p

T

AB C

d)

min

d = 1

16 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-21
SLIDE 21

Introduction to jets – Jet Algorithms in ATLAS

Sequential recombination algorithms (kt-like)

Algorithm specification: Anti-kt

di,j = min( 1 p2

T,i

, 1 p2

T,i

) ∆R2 D2 ; di,Beam = 1 p2

T,i

D : algorithm parameter Iterate:

1

For every pair of objects i, j calculate dmin = min(di,j, di,beam)

2

If dmin = di,j recombine objects Else i is a jet, remove it from list a Recombination starts from hard objects

aATLAS default: inclusive algorithm

Idea:

y p

T

A B C min A,B

a)

y p

T

AB C

b)

y p

T

AB C

c)

jet jet d = d

min AB,Beam

d = d

min C, Beam

d = d y p

T

AB C

d)

17 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-22
SLIDE 22

Introduction to jets – Jet Algorithms in ATLAS

Sequential recombination algorithms (kt-like)

Algorithm specification: Anti-kt

di,j = min( 1 p2

T,i

, 1 p2

T,i

) ∆R2 D2 ; di,Beam = 1 p2

T,i

D : algorithm parameter Iterate:

1

For every pair of objects i, j calculate dmin = min(di,j, di,beam)

2

If dmin = di,j recombine objects Else i is a jet, remove it from list a Recombination starts from hard objects

aATLAS default: inclusive algorithm

Is it safe? Yes, by construction:

Collinear, infrared safe soft particles recombined

17 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-23
SLIDE 23

Introduction to jets – Jet Algorithms in ATLAS

Sequential recombination algorithms (kt-like)

Algorithm specification: Anti-kt

di,j = min( 1 p2

T,i

, 1 p2

T,i

) ∆R2 D2 ; di,Beam = 1 p2

T,i

D : algorithm parameter Iterate:

1

For every pair of objects i, j calculate dmin = min(di,j, di,beam)

2

If dmin = di,j recombine objects Else i is a jet, remove it from list a Recombination starts from hard objects

aATLAS default: inclusive algorithm

Is it fast enough?

Time needed to cluster N particles vs N

Yes

17 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-24
SLIDE 24

Introduction to jets – Jet Algorithms in ATLAS

What jet algorithm and parameters?

Decision: choice of jet algorithm distance parameter (R) “It’s all fun and games until someone loses a hard constituent”

Example figures from original jetography paper arXiv 0810.1304: Quantifying the performance of jets, G. Salam, J. Rojo, M. Cacciari

Advantages of wider distance parameters (large-R):

Captures more QCD radiation:

→ Smaller non-perturbative corrections when comparing data to theory → Better mass resolution for dijet resonances

Dijet mass for resonance decaying into two gluons: improvement in resolution when increasing radius

Disdvantages of wider distance parameters (wider jets):

Captures more of anything else:

→ extra energy not from hard scattering (calorimeter noise, other pp collisions)

Dijet mass for resonance decaying into two gluons, large-radius: deterioration in resolution when increasing pile-up as in left to right plot with large kinematic boost, decay products

  • f heavy objects more collimated

...can we use this to our advantage?

18 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-25
SLIDE 25

Probing the substructure of jets

[ATLAS arXiv 1306.4945, Submitted to JHEP, JetETMiss WG public results]

19 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-26
SLIDE 26

Jet substructure – Introduction

Jet substructure

When to make fat jets:

When more objects (e.g. from a decay) are collimated due to kinematic boost: collect everything in a large-R (fat) jet probe substructure of this large-R jet (e.g. sub-jets)

How to use fat jets:

exploit jet grooming techniques to:

separate QCD jets from jets from boosted

  • bjects decays (background rejection)

make jets more resilient to radiation/pile-up

use jet mass as a handle for mass of heavy

  • bject (e.g. W, or top)

Example: boosted top candidate

[ATLAS-CONF-2011-073]

20 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-27
SLIDE 27

Jet substructure – Introduction

Jet substructure is an active field...

From G. Salam’s closing talk at BOOST2012

21 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-28
SLIDE 28

Jet substructure – Introduction

A famous substructure technique: mass-drop filtering [arXiv 0802.2470]

1

Find Cambridge/Aachen R=1.2 jets

2

Undo last step of jet algorithm and obtain two proto-jets (j1, j2)

3

Only keep C/A jets where:

significant difference between original jet and j1: mj1/mC/A jet < µfrac symmetric splitting between j1, j2: y =

min[(pj1

T )2, pj2 T )2]2

mC/Ajet ∆R2

j1,j2 > ycut 4

Recluster constituents of the jet using C/A with distance parameter=Rfilt,

  • nly keep three hardest subjets

22 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-29
SLIDE 29

Jet substructure – Introduction

A famous substructure technique: mass-drop filtering [arXiv 0802.2470]

It could be useful for Higgs decay in b¯ b (overwhelming background): Frequently Asked Questions

Is it really useful for boosted Higgs? We’ll know at the LHC @ 14 TeV Is it useful for ATLAS analyses? Yes, we’ll see this later

23 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-30
SLIDE 30

Jet substructure – Jet substructure performance

Substructure techniques in presence of pile-up

Original aim of jet filtering algorithms [arXiv 0802.2470]:

“filter away UE contamination while retaining hard perturbative radiation from the Higgs decay products” Impact of pile-up for anti-kT jets as a function of R

pv

N 1 2 3 4 5 6 Mean Jet Mass [GeV] / 1 PV 20 40 60 80 100 120 140 160 ATLAS

> 300 GeV, |y| < 2

T

jets, p

t

anti-k

0.1 ( = 3.0

PV

R=1.0: d m / d N 0.1 ( = 0.7

PV

R=0.6: d m / d N 0.1 ( = 0.2

PV

R=0.4: d m / d N

  • 1

L = 35 pb

Impact of pile-up for C/A jets R=1.2, before and after filtering

PV

N 1 2 3 4 5 6 7 8 9 Mean Jet Mass [GeV] / 1 PV 80 100 120 140 160 180 200 220 240 260

Before Splitting/Filtering After Splitting/Filtering After Splitting Only Cambridge-Aachen R=1.2 jets > 0.3

qq

Split/Filtered with R > 300 GeV, |y| < 2

T

p 0.3 GeV ( = 2.9

PV

dN dm 0.1 GeV ( = 4.2

PV

dN dm 0.2 GeV ( = 0.1

PV

dN dm

ATLAS

  • 1

L = 35 pb

Technique can be employed to reduce impact of pile-up

24 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-31
SLIDE 31

Jet substructure – Jet substructure performance

Substructure techniques in presence of more pile-up

High-luminosity LHC (14 TeV, after Run-II): number of additional interactions (µ) could go up to 140 and more ⇒ will jet substructure techniques still work?

Simulated impact of pile-up on QCD jet mass for R=1.0 anti-kT jets Simulated impact of pile-up on Z′ → t¯ t jet mass for R=1.0 anti-kT jets

Need both trimming and pile-up correction, but it will work!

25 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-32
SLIDE 32

Jet substructure – Jet substructure performance

Jet mass measurements

Mass of single fat, groomed jet: handle on mass of heavy boosted objects ⇒ a well known standard candle can be used to set mass scale in data

Mass distribution for C/A split/filtered jets in W → lν, with pT,W > 200 GeV Fit to the W mass distribution in 2011 data

26 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-33
SLIDE 33

Measuring the performance of jets

[ATLAS arXiv 1112.6426 and EPJC, JetETMiss WG public results]

27 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-34
SLIDE 34

Jet performance

Recap: jet reconstruction in ATLAS - jet finding

Energy deposits in calorimeters

⇒

Jet reconstruction jet finding calibration

⇒

{ }

p

i

{ }

jk

jet algorithm particles 4-momenta T

  • poclusters

jets

Many alternative jet finders can be found (e.g. kT family/Cambridge, SISCone, ...) Group objects in jet with jet algorithm

→ ATLAS default: Anti-kT

Aggregative algorithm, combines pairs of constituents sequentially Combination depends on jet pT , angular distance in (η, φ) Algorithm clusters highest energy constituents first High pT Anti-kT jets have regular shapes, stable under pile-up

28 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-35
SLIDE 35

Jet performance – Jet calibration

Jet reconstruction: calibration

Energy deposits in calorimeters

⇒

Jet reconstruction jet finding calibration

⇒

Measure energy from readout signal

→ EM / hadronic calibration

to electromagnetic scale

{ }

ADC

{ }

E

{ }

E

EM JES

EM calibration to jet energy scale HAD calibration

{ }

E

EM

Calorimeter jet response corrected for:

Non-compensating calorimeters Inactive material Out-of-cone effects

Further calibration steps:

pile-up correction to remove extra energy from multiple interactions correction based on in-situ balance techniques (e.g. γ+jet)

29 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-36
SLIDE 36

Jet performance – Jet calibration

Jet finding and calibration in ATLAS

jet constituents jets Local cluster weighting Calorimeter clusters (LCW scale) Calorimeter clusters (EM scale) Jet finding Calorimeter jets (LCW scale) Jet finding Calorimeter jets (EM scale) Tracks Track jets Simulated particles Particle jets (aka truth jets)

Calibrates clusters based on cluster properties related to shower development

Jet finding Jet finding

Calorimeter jets (EM or LCW scale) Pile-up offset correction Origin correction Energy & ! calibration Residual in situ calibration Calorimeter jets (EM+JES or LCW+JES scale)

Changes the jet direction to point to the primary vertex. Does not affect the energy. Calibrates the jet energy and pseudorapidity to the particle jet scale. Derived from MC. Residual calibration derived using in situ measurements. Derived in data and MC. Applied only to data. Corrects for the energy

  • ffset introduced by pile-up.

Depends on µ and NPV. Derived from MC.

Note: origin correction not applied in 2012

30 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-37
SLIDE 37

Jet performance – Jet energy scale uncertainty

Jet energy scale uncertainty in ATLAS

Estimate JES uncertainties using:

In-situ techniques (γ − jet, Z − jet, multijet balance, track vs calorimeter jets) single particle uncertainties from test beam convoluted to jets (high-pT ) pT balance in dijet events (forward JES uncertainty) Different MC generators (jet flavor and topology uncertainty)

Comparison: before collisions → 2011 (with in-situ correction)

<6.5 → 1% for central jets, pT =200 GeV <10 → 9% for more forward jets

[GeV]

jet T

p 20 30 40

2

10

2

10 × 2

3

10

3

10 × 2 Fractional JES uncertainty 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 ATLAS Preliminary

= 0.4, EM+JES R

t

Anti-k = 0.5 η in situ 2011 2010 Before collisions Baseline JES uncertainties

Overall JES uncertainty for 2012

As low as 1% for central jets, pT =250 GeV

31 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-38
SLIDE 38

Jet performance – Jet energy scale uncertainty

Jet energy scale uncertainty correlations

Treatment of correlated experimental uncertainties:

ATLAS has o(60) jet energy scale nuisance parameters:

set of uncertainty sources, each correlated bin-to-bin, uncorrelated among themselves

Propagation through analysis allows quantitative theory comparisons, meaningful inclusion in PDF fits, combinations with other experiments

Components:

pile-up uncertainties uncertainty sources from in-situ techniques: systematic and statistical high-pT (single particle) uncertainties flavor and topology uncertainties b−jet uncertainties

In-situ correlation matrix (2011): How can an analysis propagate 60 uncertainty sources separately? (they could, it

just takes a long time...) ⇒ Nuisance parameter reduction technique:

Fewer nuisance parameters, still retaining information on correlations and category for combination with other experiments (e.g. uncertainty from detector

effects, MC modeling effects...)

32 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-39
SLIDE 39

Jet performance – Jet resolution

Jet energy resolution

Jet energy resolution: reflects intrinsic fluctuations of reconstructed jet energy from true

jet energy

Two independent in-situ techniques to estimate JER and compare to MC Up to 30% improvement if using the more refined calibration technique

33 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-40
SLIDE 40

Measuring the Standard Model with jets

34 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-41
SLIDE 41

Standard Model jet results – Jet triggers

Jet triggers

The ATLAS trigger system

3-tier system (Level-1, Level-2, Event Filter) Reduces data intake from ≈ o(10) MHz to ≈ 300 Hz Jet triggers: allow for rejection of fakes at L2, anti-kT jets at the event filter

ATLAS jet triggers (Summer 2011):

1

Minimum Bias Scintillators (MBTS)

2

Single-jet triggers (central and forward)

3

Multijet triggers

4

Topology based triggers

5

Combination triggers

[ATL-DAQ-PROC-2011-034]

35 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-42
SLIDE 42

Standard Model jet results – Jet triggers

Jet triggers

The ATLAS trigger system

3-tier system (Level-1, Level-2, Event Filter) Reduces data intake from ≈ o(10) MHz to ≈ 300 Hz Jet triggers: allow for rejection of fakes at L2, anti-kT jets at the event filter

|y| 2.8 3.0 3.2 3.4 3.6 Trigger Efficiency 0.2 0.4 0.6 0.8 1 1.2 ATLAS

  • 1

dt = 34 pb L

∫

= 7 TeV, s = 0.6 R jets,

t

anti-k pT > 60 GeV > 10 GeV

EM T

E central trigger, L1 > 10 GeV

EM T

E forward trigger, L1 central OR forward

(a) Trigger combination for 2010 inclu- sive jet cross-section in the transition re- gion between two trigger systems (b) Example of jet trigger efficiency curves for Level-2 in 2012

35 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-43
SLIDE 43

Standard Model jet results – Measurement of jet properties

Low-momentum jets and non-perturbative QCD

Measure properties of low-momentum jets using jets reconstructed from tracks : probe non-perturbative QCD from minimum bias to jet structure at higher pT

[ATLAS arXiv 1107.3311, PRD ]

b/GeV] µ dy [

T

/dp σ

2

d

  • 3

10

  • 2

10

  • 1

10 1 10

2

10

3

10

= 7 TeV s Data 2010 Pythia6 AMBT1 Herwig++ 2.4.2 Herwig++ 2.5.1 UE7 Pythia 8.145 4C Pythia6 Perugia 0 Pythia6 Perugia 2010 Pythia6 DW Phojet + Pythia6

R = 0.6

ATLAS

[GeV]

T

Charged Jet p 5 6 7 8 10 20 30 40

2

10

  • 0.4
  • 0.2

0.2 0.4 0.6

Data MC - Data

jet ch

N 5 10 15 20 25

  • 1

1

Pythia6 Perugia 0 Pythia6 Perugia HARD Pythia6 Perugia SOFT Pythia6 DW Phojet + Pythia6

  • 1

1

  • 1

1

  • 1

1

  • 1

1

Data MC - Data

4 - 6 GeV 6 - 10 GeV 10 - 15 GeV

R = 0.4

ATLAS

15 - 24 GeV 24 - 40 GeV

Track-jet cross-section and charged particle multiplicity, anti-kT R=0.6

36 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-44
SLIDE 44

Standard Model jet results – Measurement of jet properties

Jet fragmentation and shape

Probe internal jet structure with measurements of charged particles inside the jet: jet fragmentation function and transverse jet profile

[ATLAS arXiv 1109.5816, EPJC ]

Measurement of jet fragmentation function: probability of charged particle carrying momentum fraction z

z = pjet · pch pjet Sensitivity to:

Fragmentation models: benchmarks for simulation Non perturbative hadronisation effects

<z>

0.03 0.04 0.05 0.06 0.07 0.08

Data Pythia6 AMBT1 Pythia6 MC09 Pythia6 Perugia 2010 Herwig+Jimmy Herwig++ 2.4.2 Herwig++ 2.5.1 Sherpa Pythia8 8.145 4C

ATLAS = 7 TeV s

  • 1

L dt = 36 pb

∫

(GeV)

T jet

p 30 40 50

2

10

2

10 × 2

(MC-Data)/Data (%)

  • 15
  • 10
  • 5

5 10 15

(GeV)

T jet

p 30 40 50

2

10

2

10 × 2

(MC-Data)/Data (%)

  • 15
  • 10
  • 5

5 10 15 37 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-45
SLIDE 45

Standard Model jet results – Measurement of jet properties

Jet fragmentation and shape

Probe internal jet structure with measurements of charged particles inside the jet: jet fragmentation function and transverse jet profile

[ATLAS arXiv 1109.5816, EPJC ]

Measurement of integrated jet shape: density of ch. particles around jet axis

Jet R r ∆r

Sensitivity to:

Fragmentation models: benchmarks for simulation Non perturbative hadronisation effects

No MC model describes both jet fragmentation and jet profile

(r)

ch

ρ

10

2

10

3

10

< 500 GeV

T jet

400 GeV < p ATLAS = 7 TeV s

  • 1

L dt = 36 pb

∫

Data Pythia6 AMBT1 Pythia6 MC09 Pythia6 Perugia 2010 Herwig+Jimmy Herwig++ 2.4.2 Herwig++ 2.5.1 Sherpa Pythia8 8.145 4C

r

0.1 0.2 0.3 0.4 0.5 0.6

(MC-Data)/Data (%)

  • 20
  • 10

10 20

r

0.1 0.2 0.3 0.4 0.5 0.6

(MC-Data)/Data (%)

  • 20
  • 10

10 20 37 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-46
SLIDE 46

Standard Model jet results – Jets, dijets and multijets

Inclusive jet, dijet and multijet cross section

Jet production: dominant high pT process at LHC

Probe perturbative QCD at small distances Understand dominant background for many analyses Early testing ground for jet calibration and performance very first measurements: 17 nb−1 [ATLAS arXiv 1012.4389, EPJC ]

[ATLAS-CONF-2010-084]

Multijet cross section:

[ATLAS arXiv 1107.2092, EPJC ]

2 3 4 5 6 [pb] σ 10

2

10

3

10

4

10

5

10

6

10 2 3 4 5 6 [pb] σ 10

2

10

3

10

4

10

5

10

6

10 ATLAS

  • 1

L dt=2.4 pb

∫

R=0.4, =7 TeV)+syst. s Data ( 1.11 × ALPGEN+HERWIG AUET1 0.65 × PYTHIA AMBT1 1.22 × ALPGEN+PYTHIA MC09’ 1.06 × SHERPA

Inclusive Jet Multiplicity 2 3 4 5 6 MC/Data 0.5 1 1.5 Inclusive Jet Multiplicity 2 3 4 5 6 MC/Data 0.5 1 1.5

Dijet double-differential cross section,

[ATLAS CONF-2012-021]

[TeV]

12

m

  • 1

10 × 3 1 2 3 4 5 6 7

* [pb/TeV] y d

12

m /d σ

2

d

  • 3

10

  • 1

10 10

3

10

5

10

7

10

9

10

11

10

13

10

15

10

17

10

) 10 × * < 0.5 ( y )

2

10 × * < 1.0 ( y ≤ 0.5 )

4

10 × * < 1.5 ( y ≤ 1.0 )

6

10 × * < 2.0 ( y ≤ 1.5 )

8

10 × * < 2.5 ( y ≤ 2.0 uncertainties Systematic Non-pert. corr. × )) y* exp(0.3

T

p = µ (CT10, NLOJET++

ATLAS Preliminary

= 0.6 R jets,

t

anti-k

  • 1

dt = 4.8 fb L

∫

= 7 TeV, s 2011 Data

38 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-47
SLIDE 47

Standard Model jet results – Jets, dijets and multijets

Jet cross sections: 7 TeV and 2.76 TeV

Measure jet cross sections at two center of mass energies (7 and 2.76 TeV): exploit uncertainty correlations, use both datasets as input for PDF fits

[ATLAS arXiv 1304.4739,EPJC] Qualitative comparison of jet cross sections for various experiments Effect of 7 and 2.76 TeV fits on gluon PDF (ATLAS+HERA data only)

39 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-48
SLIDE 48

Searching for new phenomena with jets

40 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-49
SLIDE 49

Searches with jets – Dijet analysis

Overview of dijet searches

Searches in the dijet mass spectrum

Select high mjj events (mjj >1000 GeV) Fit QCD background from data using smooth function: f(x) = p1(1 − x)p2 · xp3+p4 log x, x = mjj/√s Look for discrepancies using BumpHunter [1101.0390] If no surprises, test models → set limits:

Benchmark: excited quark (q∗) production [PRD] Color octet model [JHEP] QCD

dijet mass mjj dσ dmjj

New Physics

|η1- η2| < 1.3

[F. Ruehr, LPCC Workshop on Higgs/BSM]

41 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-50
SLIDE 50

Searches with jets – Dijet analysis

Overview of dijet searches

Searches in dijet angular distributions

Select high mjj events (mjj >850 GeV) Look for excesses above QCD at high scattering angles Use Fχ(mjj) = Ncentral Ntotal to resolve evolution of angular shape in fine mass bins Use normalised χ = e|y1−y2| distribution for angular shape in wide mass bins

Benchmark: Contact Interactions Quantum Black Holes [JHEP]

dijet mass mjj Fχ(mjj)

QCD

New Physics

χ = exp |y1-y2| (1/N) dN/dχ

QCD

New Physics

[F. Ruehr, LPCC Workshop on Higgs/BSM]

41 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-51
SLIDE 51

Searches with jets – Dijet analysis

Searches in the dijet mass distribution

[Search for dijet mass resonances: ATLAS-CONF-2012-148] Look for resonances above smooth background in central dijet mass spectrum: none found Set 95% C.L. limit on σ × A for excited quark model (m(q*) < 3.84 TeV) Include model-independent limits on Gaussian resonances of varying width

2000 3000 4000 1 10

2

10

3

10

4

10

5

10

Data Background [GeV]

jj

Reconstructed m 2000 3000 4000 Events Significance

  • 2

2

ATLAS Preliminary

  • 1

= 13.0 fb dt L

∫

= 8 TeV s

[GeV]

G

Mass, m

2000 3000 4000

[pb] xA × σ 95% CL Limit on

  • 4

10

  • 3

10

  • 2

10

  • 1

10 1

G

/ m

G

σ 0.15 0.10 0.07

ATLAS Preliminary

  • 1

= 13.0 fb dt L

∫

= 8 TeV s

No evidence of new phenomena with more than half of the 8 TeV dataset Consistence with good agreement of SM measurement of mjj with QCD at 7 TeV

42 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-52
SLIDE 52

Searches with jets – Dijet analysis

Searches in dijet angular distributions

[ATLAS-CONF-2010-056]

Fχ(mjj) = N(|y ∗ | < 0.6) N(|y ∗ | < 1.7) distribution, with QCD prediction superimposed with 95% C.L. limit on Quantum Black Holes model as a function of Planck mass

[GeV]

jj

m 1000 1500 2000 2500 3000 3500 4000 4500

χ

F 0.1 0.2 0.3 0.4 0.5 0.6 ATLAS Preliminary

=7 TeV s ,

  • 1

dt = 4.8 fb L

∫

QCD Prediction Theoretical uncertainties Total Systematics data = 7.5 TeV Λ Contact Interaction: = 4.5 TeV

D

QBH (n=6) : M = 2.5 TeV

q*

q*: m Upper boundary to control region Lower boundary to search region

(a)

[GeV]

D

M 2500 3000 3500 4000 4500 5000 [pb] xA × σ

  • 4

10

  • 3

10

  • 2

10

  • 1

10 1 10

2

10

3

10

4

10

5

10 ATLAS Preliminary

) variable

jj

(m

χ

Limits set using F QBH n=2 QBH n=3 QBH n=4 Observed 95% CL upper limit Expected 95% CL upper limit 68% band 95% band QBH n=5 QBH n=6 QBH n=7 =7 TeV s ,

  • 1

dt = 4.8 fb L

∫

(b)

No evidence of new phenomena with 7 TeV dataset Consistent with good agreement of SM jet measurements with QCD

43 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-53
SLIDE 53

Searches with jets – Dijet analysis

Searches in dijet angular distributions

[ATLAS-CONF-2010-056]

The χ = exp (|y1 − y2|) distribution, with QCD prediction superimposed with summary table for 95% C.L. limits

χ 1 10 χ / d σ ) d σ (1/ 0.05 0.1 0.15 0.2 0.25 0.3 0.35

< 1200 GeV

jj

800 < m < 1600 GeV (+0.04)

jj

1200 < m < 2000 GeV (+0.08)

jj

1600 < m < 2600 GeV (+0.12)

jj

2000 < m > 2600 GeV (+0.16)

jj

m QCD Prediction Theoretical uncertainties Total Systematics = 4.0 TeV

D

QBH (n=6), M =7 TeV s ,

  • 1

dt = 4.8 fb L

∫

ATLAS Preliminary

No evidence of new phenomena with 7 TeV dataset Consistent with good agreement of SM jet measurements with QCD

43 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-54
SLIDE 54

Searches with jets – Photon+jet analysis

Searches in the γ+jet invariant mass spectrum

Select central high pT γ-jet events (pT,γ, pT,jet >125 GeV) Build γ − jet invariant mass

Reject background using calorimeter isolation/topology

Fit background from data using smooth function, look for discrepancies using BumpHunter [1101.0390] No evidence of new phenomena in entire 8 TeV dataset

[ATLAS arXiv 1309.3230, sub. to PRD]

No surprises → set limits on benchmark models:

Excited quarks (q*) Quantum Black Holes Hypothetical Gaussian γ−jet decay signal (mass mG, width σG )

44 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-55
SLIDE 55

Searches with jets – Mono-X analyses for dark matter

Searches for dark matter in mono-X+MET

From cosmological and astroparticle experiment observations: ≈ 95% of the universe is (directly or indirectly) evident but unexplained: dark matter and dark energy

[arXiv 1305.1605]

Synergy needed with other experiments for dark matter detection in space and in labs

45 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-56
SLIDE 56

Searches with jets – Mono-X analyses for dark matter

Searches for dark matter in mono-X+MET

From cosmological and astroparticle experiment observations: ≈ 95% of the universe is (directly or indirectly) evident but unexplained: dark matter and dark energy

χ ¯ χ q q g,γ,Z,or W=jet!

Effective Field Theory

LHC experiments have a shot at finding a particle candidate for dark matter: dark matter interacts gravitationally ⇒ could it interact weakly?

Specific, UV-complete theories: e.g. SUSY, with Lightest Supersymmetric Particle as DM candidate → optimise sensitivity for certain models Simplified models: e.g. effective theory encompassing interaction between SM and DM particles → less sensitive but more generic

45 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-57
SLIDE 57

Searches with jets – Mono-X analyses for dark matter

Searches in the monojet final state

[Search for new physics in monojet: ATLAS CONF-2012-147] Select events with large jet pT and missing transverse momentum Background estimation: use transfer factors from control regions in data Counting experiment: hope to observe excess of events above jet pT and missing transverse momentum thresholds Set model-independent limits (σ × A), limits on Large Extra Dimensions, WIMPs, Gravitinos

No significant excess over background in 10 fb−1

Data, background and example signals in

  • ne of the signal regions

[Events/GeV]

T

dN/dp

  • 1

10 1 10

2

10

3

10

4

10

5

10

data 2012 Total BG ) + jets ν ν → Z ( ) + jets ν l → W ( Multi-jet Non-collision BG ll ) + jets → Z ( Dibosons + single top t t =3 TeV (x5)

D

ADD n=2, M =670GeV (x5)

*

D5 M=80GeV, M eV (x5)

  • 4

=10

G ~

=1TeV, M

g ~ / q ~

, M g ~ / q ~ + G ~

  • 1

Ldt=10.5fb

∫

= 8 TeV s ATLAS Preliminary [Events/GeV]

T

dN/dp

  • 1

10 1 10

2

10

3

10

4

10

5

10 jet1 [GeV]

T

p 200 400 600 800 1000 1200

Data / BG

0.5 1 1.5

Errors on plot are statistical only 8 TeV analysis limited by modelling uncertainties

46 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-58
SLIDE 58

Searches with jets – Mono-X analyses for dark matter

Searches in the monojet final state

[Search for new physics in monojet: ATLAS CONF-2012-147] Select events with large jet pT and missing transverse momentum Background estimation: use transfer factors from control regions in data Counting experiment: hope to observe excess of events above jet pT and missing transverse momentum thresholds Set model-independent limits (σ × A), limits on Large Extra Dimensions, WIMPs, Gravitinos

No significant excess over background in 10 fb−1

Limits on WIMP scalar operator D11

[GeV]

χ

WIMP mass m

2

10

3

10 [GeV]

*

Suppression scale M 100 150 200 250 300 350 400 450 500

, SR3, 90%CL Operator D11 )

exp

σ 2 ± 1 ± Expected limit ( )

theory

σ 1 ± Observed limit ( Thermal relic

Preliminary ATLAS

=8 TeV s

  • 1

Ldt = 10.5 fb

∫

not valid effective theory

Caveat: validity of effective theories at colliders → Theory/experiment collaborations to ensure complementarity of DM searches Higgs entering searches: reinterpretation in terms of H→ invisible BR (and vice-versa: H→ inv. reinterpreted as limits on WIMPs)

46 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-59
SLIDE 59

Searches with jets – Mono-X analyses for dark matter

Searches in mono-W final state

If dark matter has opposite-sign couplings to up and down quarks → preferential radiation of W boson

[Search for dark matter in mono-W: ATLAS arXiv-1309.4017, Submitted (yesterday) to PRL] Select events with C/A split/filtered jet, with mass around W peak (hadronic W) + missing transverse momentum

[GeV]

jet

m

50 60 70 80 90 100 110 120 Events / 10 GeV 20 40 60 80 100 120 140 160 180

Data Top )+jet τ / µ W(e/ uncertainty

ATLAS = 8 TeV s

  • 1

20.3 fb > 250 GeV

miss T

top CR: E

Main background estimation technique: similar to monojet (transfer factors) Look for excess over number of estimated events: no excess found in whole 8 TeV dataset Data, background and hypothetical signal

Events / 10 GeV 50 100 150 200 250 Data )+jet ν ν Z( )+jet τ / µ W/Z(e/ Top Diboson uncertainty D5(u=d) x100 D5(u=-d) x1 ATLAS = 8 TeV s

  • 1

20.3 fb > 350 GeV

miss T

SR: E

[GeV]

jet

m

50 60 70 80 90 100 110 120 Events / 10 GeV 5 10 15 20 25 30 35 D5(u=d) x20 D5(u=-d) x0.2 > 500 GeV

miss T

SR: E 47 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-60
SLIDE 60

Searches with jets – Mono-X analyses for dark matter

Searches in mono-W final state

If dark matter has opposite-sign couplings to up and down quarks → preferential radiation of W boson

[Search for dark matter in mono-W: ATLAS arXiv-1309.4017, Submitted (yesterday) to PRL] Select events with C/A split/filtered jet, with mass around W peak (hadronic W) + missing transverse momentum

[GeV]

jet

m

50 60 70 80 90 100 110 120 Events / 10 GeV 20 40 60 80 100 120 140 160 180

Data Top )+jet τ / µ W(e/ uncertainty

ATLAS = 8 TeV s

  • 1

20.3 fb > 250 GeV

miss T

top CR: E

Main background estimation technique: similar to monojet (transfer factors) Look for excess over number of estimated events: no excess found in whole 8 TeV dataset Limit on dark matter-nucleon cross section, compared to other experiments

[GeV]

χ

m 1 10

2

10

3

10

  • 46

10

  • 44

10

  • 42

10

  • 40

10

  • 38

10

  • 36

10

SIMPLE 2011

  • W

+

IceCube W b IceCube b COUPP 2012 PICASSO 2012 D9:obs ) χ χ D9: ATLAS 7TeV j( = 8 TeV s

  • 1

20.3 fb

spin-dependent ATLAS [GeV]

χ

m 1 10

2

10

3

10 ]

2

  • N cross-section [cm

χ

  • 46

10

  • 44

10

  • 42

10

  • 40

10

  • 38

10

  • 36

10

D5(u=-d):obs D5(u=d):obs ) χ χ D5:ATLAS 7TeV j( COUPP 2012 CoGeNT 2010 XENON100 2012 CDMS low-energy

spin-independent 90% CL

Also available: limit on Higgs to invisible branching ratio (σinv/σtot SM 1.6)

47 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-61
SLIDE 61

Conclusions

Conclusions and outlook

A wealth of jet results produced by ATLAS in 2011 7 and 8 TeV dataset: more precision measurements and searches in the pipeline

Jet algorithms and performance:

No one fits all jet algorithm: complexity of jets can be exploited (e.g. jet substructure) Good understanding of the jet energy scale and resolution in ATLAS data throughout the LHC Run-I

Standard Model jet measurements:

Good agreement of data and pQCD Effort ongoing to deliver ATLAS data for constraining PDFs/theory/MC models

Searches with jets:

No evidence of new phenomena in jet final states Exclusion limits set on exotic models

48 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar
slide-62
SLIDE 62

Conclusions

Conclusions and outlook

A wealth of jet results produced by ATLAS in 2011 7 and 8 TeV dataset: Expect much more with the 13 TeV data: let’s prepare for the excitement of dijet searches in 2015! Thanks for your attention!

48 / 48

  • C. Doglioni - 17/09/2013 - UVa Seminar