Hard QCD Results with Jets @ LHC (ATLAS + CMS) Hadron Collider Physics Symposium 2011 Sven Menke, MPP M ¨ unchen, 16. Nov. 2011, Paris, France on behalf of the ATLAS and CMS collaborations Introduction • ATLAS+CMS • jet reconstruction and calibration Cross sections • inclusive jet cross sections • di-jet event studies Jet properties • sub structure • jet mass Conclusions Highest Energetic Jet in ATLAS from 2010 @ √ s = 7 TeV with E j = 3.37 TeV
26
Embed
Hard QCD Results with Jets @ LHC (ATLAS + CMS) · Total Delivered: 48.1 pb1 Total Recorded: 45.0 pb1 LHC: pp Collisions @ p s = 7 TeV since March 2010 2011 ATLAS/CMS recorded integrated
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Hard QCD Results with Jets @ LHC(ATLAS + CMS)
Hadron Collider Physics Symposium 2011 Sven Menke, MPP Munchen, 16. Nov. 2011, Paris, Franceon behalf of the ATLAS and CMS collaborations
� Introduction• ATLAS+CMS
• jetreconstructionand calibration
� Crosssections• inclusive jet
cross sections
• di-jet eventstudies
� Jetproperties• sub structure
• jet mass
� ConclusionsHighest Energetic Jet in ATLAS from 2010 @
� Traditionally all calorimeter cells are grouped in fixed size towers with∆φ×∆η = 0.1× 0.1 including all em and hadronic layers per grid point
� ATLAS and CMS use refined reconstruction methods to feed jetalgorithms
� CMS uses Particle Flow (PF)• PF reconstructs leptons, photons, charged and neutral hadrons by
linking tracks to ECAL and HCAL clusters• energy of each particle measured by all subsystems• ECAL and HCAL energy is calibrated to respective particle level
(electromagnetic or hadronic)
� ATLAS uses TopoClusters• topological 3D clustering over all 190 k calorimeter cells• individual energy blobs separated by local maxima correspond to
particle level• in 2010 em-scale clusters were used as input to jets• jet-level corrections correct for non-compensation, PileUp, dead
material
S. Menke, MPP Munchen � Hard QCD Results with Jets @ LHC � HCP 2011, 16. Nov. 2011, Paris, France 6
� Modern standardized jet algorithms like SISCone (JHEP 0705 (2007) 086), Kt(Nucl. Phys. B 406 (1993) 187, Phys. Rev. D 48 (1993) 3160), and AntiKt (JHEP 0804 (2008) 063)have been evaluated by ATLAS and CMS• these jets are collinear and infrared safe• are available in a standard C++ library (FastJet by Matteo Cacciari, Gavin Salam
and Gregory Soyez)• are seedless and iterative
� AntiKt in inclusive mode was found to be the most useful algorithm forATLAS and CMS• AntiKt combines like Kt pairs of objects based on a
min(
piT
2x, pj
T2x)
scaled distance metric ∆R2ij /R
2 in y − φ-space
• Kt uses x = 1 and treats objects with the smallest pT first• AntiKt uses x = −1 and treats objects with the largest pT first• the net result is that AntiKt-jets are much more regular shaped in
y − φ space and don’t suffer from the “vacuum cleaner” effect like Ktmaking them easier to calibrate
• CMS uses R = 0.5 (incl.) and R = 0.7 (di-jet), while ATLAS usesR = 0.4 and R = 0.6 (both)
S. Menke, MPP Munchen � Hard QCD Results with Jets @ LHC � HCP 2011, 16. Nov. 2011, Paris, France 7
� most analyses on 2010 data used simpleEM+JES scheme• topo clusters on EM-scale as input to jet algorithms• MC based correction function f(p⊥, |η|) to restore jet p⊥ to
stable hadron level• top plot shows correction as function of p⊥ for 3 y ranges• middle plot shows systematic uncertainties on jet energy scale
(JES) for 0.3 < |y| < 0.8
� more sophisticated approaches exist• global cell weighting (GCW) applies energy-density
dependent weights to all calorimeter cells• local hadron calibration (LCW) classifies and calibrates topo
clusters as em or hadronic• global sequential calibration (GSW) modifies EM+JES by
jets-shape based correction factors• bottom plot compares light-quark – gluon-jet-response for
different calibration schemes
� LCW is used in ATLAS for missing transverseenergy and in jets for 2011
� Jet mass and substructure studies reportedhere also used this already for 2010 data
[GeV]jet
Tp
20 30 210 210×23
103
10×2
Avera
ge J
ES
corr
ection
1
1.2
1.4
1.6
1.8
2| < 0.8η0.3 < |
| < 2.8η2.1 < |
| < 4.4η3.6 < |
= 0.6, EM+JESR t
Antik
ATLAS Preliminary
[GeV]jet
Tp
30 40 210 210×23
103
10×2
Fra
ctio
na
l JE
S s
yste
ma
tic u
nce
rta
inty
0
0.02
0.04
0.06
0.08
0.1
0.12
ATLAS Preliminary
| < 0.8, Data 2010 + Monte Carlo QCD jetsη | ≤=0.6, EM+JES, 0.3 R t
Antik
ALPGEN + Herwig + Jimmy Noise Thresholds
JES calibration nonclosure PYTHIA Perugia2010
Single particle (calorimeter) Additional dead material
Total JES uncertainty
[GeV]true
Tp
30 40 100 200 300
Lig
ht
qu
ark
G
luo
n J
et
Re
sp
on
se
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1ATLAS Preliminary
SimulationPYTHIA AMBT1
EM+JES CalibrationGCW Calibration
LCW CalibrationGS Calibration
|<2.8η |≤2.1 R=0.6 Jets
tantik
S. Menke, MPP Munchen � Hard QCD Results with Jets @ LHC � HCP 2011, 16. Nov. 2011, Paris, France 9
� most analyses use jets with particle flow (PF)objects as input
� jet energy scale uncertainties are evaluated forthree types of jets in CMS• calorimeter only jets• jet-plus-track jets (calorimeter jets with corrections based on
associated tracks)• particle flow jets (all subsystems are used to reconstruct individual
stable particles)
� in all three cases the jet energy is corrected by aMC based correction function in jet p⊥ and η• top plot shows jet energy correction factors for central jets vs. p⊥
� scale and uncertainty are validated with di-jetevents (p⊥ balance for relative JES) and γ-jetevents (MPF and p⊥ balance for absolute JES)• middle plot shows absolute JES uncertainty for particle-flow jets
vs. p⊥• bottom plot shows total JES uncertainty for all three jet types vs. p⊥
for central jets
(GeV)T
Jet p30 100 200 1000
Je
t E
ne
rgy C
orr
ectio
n F
acto
r
0.8
1
1.2
1.4
1.6
1.8
2
2.2
Calorimeter jets
Jet-Plus-Track jets
Particle Flow jets
= 7 TeVs R = 0.5
Tanti-k
= 0.0η
CMS Preliminary
Je
t E
ne
rgy C
orr
ectio
n F
acto
r
Je
t E
ne
rgy C
orr
ectio
n F
acto
r
TT
(GeV)T
p20 30 100 200 1000 2000
Absolu
te s
cale
uncert
ain
ty [%
]
0
1
2
3
4
5
6
7
8
9
10Total uncert.Total MPF
Photon scaleExtrapolationOffset (+1PU)
Residuals
particle flow jets
0.5 & 0.7T
Anti-k
= 7 TeVs-1CMS preliminary, 2.9 pb
(GeV)T
p20 30 100 200 1000 2000
Absolu
te s
cale
uncert
ain
ty [%
]
0
1
2
3
4
5
6
7
8
9
10
(GeV)T
Jet p20 30 100 200 1000
To
tal JE
S U
nce
rta
inty
[%
]
0
5
10
Calorimeter jets
Jet-Plus-Track jets
Particle Flow jets = 7 TeVs
R = 0.5T
anti-k
| = 0.0η|
CMS Preliminary
To
tal JE
S U
nce
rta
inty
[%
]
S. Menke, MPP Munchen � Hard QCD Results with Jets @ LHC � HCP 2011, 16. Nov. 2011, Paris, France 10
� measured spectrum is corrected bin-by-bin for migration effects in p⊥• by Gaussian smearing with p⊥ resolution of modified power-law
spectrum obtained by fit to data (CMS)• by correction factors obtained from full detector simulations including
also detector inefficiencies (ATLAS)
� NLO pQCD predictions• NLOJet++ 4.1.2 with CTEQ 6.6 NLO PDFs as baseline (ATLAS)• NLOJet++ 2.0.1 with average of CT10, MSTW2008NLO, NNPDF2.0 NLO
PDFs as baseline (CMS)
� NLO predictions on parton-level are corrected bin-by-bin for non-perteffects due to hadronisation and underlying event• from ratio of leading log generator spectrum with and without
hadronisation and underlying event (ATLAS+CMS)
S. Menke, MPP Munchen � Hard QCD Results with Jets @ LHC � HCP 2011, 16. Nov. 2011, Paris, France 11
Cross sections � di-jet events� double differential cross-section in |y|max and m12� both ATLAS and CMS use bin-by-bin corrections in the observed spectra derived from simulations� deviations from the QCD would indicate new physics
� compositeness, excited quarks� good agreement with NLO QCD predictions is observed
[TeV]12
m
110 110×2 1 2 3 4 5
[p
b/T
eV
]m
ax
|y
d|
12
m/d
σ2
d
210
1
210
410
610
810
1010
1210
1410
1610
1810
2010
2110Systematic uncertainties
Nonpert. corr.×NLO pQCD (CTEQ 6.6)
)8
10× < 2.8 (max|y2.1 < |
)6
10× < 2.1 (max|y1.2 < |
)4 10× < 1.2 (max|y0.8 < |
)2 10× < 0.8 (max|y0.3 < |
)0
10× < 0.3 (max|y |
ATLAS Preliminary
= 0.4R jets, tantik
1 dt = 37 pbL∫ = 7 TeV, s
(TeV)JJ
M0.2 0.3 1 2 3 4
(pb/T
eV
)m
ax
d|y
|JJ
/dM
σ2
d
-310
210
710
1210
1710 < 0.5
max|y|
)1 10× < 1.0 (max
0.5 < |y|
)2 10× < 1.5 (max
1.0 < |y|
)3
10× < 2.0 (max
1.5 < |y|
)4 10× < 2.5 (max
2.0 < |y|
Non Pert. Corr.⊗pQCD at NLO
PDF4LHCave
T = p
Rµ =
Fµ
-1 = 36 pb
intL
= 7 TeVs
R = 0.7T
anti-k
CMS
R = 0.4 (ATLAS) R = 0.7 (CMS)� dominant exp. systematic uncertainties stem from the Jet Energy Scale uncertainty (ATLAS+CMS)
• ∼ 15− 30% (ATLAS)• ∼ 15% at low mass;∼ 60% at high mass (CMS)
S. Menke, MPP Munchen � Hard QCD Results with Jets @ LHC � HCP 2011, 16. Nov. 2011, Paris, France 16
Di-jet events � with a veto on central jet activity
� Look at the activity inside therapidity gap between a di-jetsystem defined by• either the two leading jets in p⊥• or the two jets with the largest rapidity gap
∆y• gap fraction is the fraction of events without
additional jet activity in the rapidity intervalbetween the jets
• the veto scale is Q0 = 20 GeV or larger tostay away from ΛQCD
� plot to the right shows gapfraction for leading jets in p⊥ asfunction of ∆y for various p⊥ranges (ATLAS)
� Powheg+Pythia describes databest; HEJ deviates at high p⊥
< 270 GeV (+3)T
p ≤240
< 240 GeV (+2.5)T
p ≤210
< 210 GeV (+2)T
p ≤180
< 180 GeV (+1.5)T
p ≤150
< 150 GeV (+1)T
p ≤120
< 120 GeV (+0.5)T
p ≤90
< 90 GeV (+0)T
p ≤70
Data 2010
HEJ (parton level)
POWHEG + PYTHIA
POWHEG + HERWIG
dijet selectionT
Leading p
= 20 GeV0
Q
ATLAS
y∆
0 1 2 3 4 5 6
Ga
p f
ractio
n
0
1
2
3
4
5
S. Menke, MPP Munchen � Hard QCD Results with Jets @ LHC � HCP 2011, 16. Nov. 2011, Paris, France 19
� Calorimeter showers are grouped by a topological clustering algorithminto 3D energy blobs
� Noise related threshold control seeding and growth of the clusters
� Proto-clusters are split around local energy maxima
� Example below shows a typical simulated QCD event in forward region(just the first layer) with different noise thresholds and final clusteringand splitting applied
� The algorithm suppresses calorimeter noise and leads to dynamicallysized clusters corresponding on average to 1.6 final state particles