Measurement of Jet Quenching in Pb-Pb sNN = …...2016/04/21 · Hiroki Yokoyama : 2nd year Thesis...
Transcript of Measurement of Jet Quenching in Pb-Pb sNN = …...2016/04/21 · Hiroki Yokoyama : 2nd year Thesis...
Measurement of Jet Quenching in Pb-Pb Collisions at √sNN = 5.02 TeV in the ALICE Experiment at the LHCHiroki Yokoyama
LPSC, Université Grenoble-Alpes, CNRS/IN2P3 University of Tsukuba, Japan
21/04/2016 2nd year Thesis Monitoring Seminar
1
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
OutlineQuark-Gluon Plasma and Jet Quenching
Quark-Gluon PlasmaJets in Heavy Ion Collisions
The ALICE Experiment at the LHC Development and commissioning of the ALICE Calorimeter L1 Trigger system
L1-Jet Trigger Algorithmbackground subtraction methodTrigger Performance
Inclusive jet measurement with √sNN = 5.02 TeV Pb-Pb collisions Jet ReconstructionUnderlying EventUnfolding Technique
Summary
2
Quark-Gluon Plasma (QGP) and Jet Quenching
3
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
A brief journey to the beginning of the Universe
Quark-Gluon Plasma (QGP) Hot & dense color thermalized QCD matter prevailing at the early Universe ~1μs after big bangDeconfined state of quarks and gluonsTheoretically inferred through lattice gauge simulations of QCD
What is QGP ?
How to create ?‘Little Bang’
high-energy head-on nucleus-nucleus collisions at particle acceleratorsRecreate QGP droplets for a brief period of time to quantitatively map out the QCD phase diagram
4
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Collimated spray of hadrons produced by the hard scattering of partons at the initial stage of the collisionhigh-Q2 process, pT ≳ 20 GeV
Jets in HI Collisions ( Hard Probes of the QGP )
Attenuation or disappearance of observed Jets in Pb-Pb
due to partons' energy loss in the QGPevaluation of the degree of the attenuation gives us QGP properties
What’s a Jet ? Jet Quenching
5
Why Jets ?The QGP lifetime is so short (~10-23s) that characterization by external probes is ruled out
self-produced probe Occur at early stage ~1/Q
probe the entire medium evolution Production rate calculable within pQCD
well calibrated probe Large cross-section at the LHC
copious production Reconstructed Jet enables to access
4-momentum of original parton jet structure (energy re-distribution)
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
ALICE Jet Quenching Measurements in Pb-PbNuclear modification factor : RAA
if RAA = 1, NO modification of the matter
)c (GeV/Tp
0 2 4 6 8 10 12 14 16 18 20
AAR
0.2
0.4
0.6
0.8
1=2.76 TeVNNsALICE 0-5% Pb-Pb
-π++π-+K+K
pp + Charged
ALI−DER−86149)c (GeV/
T,jetp
0 50 100
AAR
0
0.2
0.4
0.6
0.8
1
1.2 = 2.76 TeVNNsALICE Pb-Pb
Data 0 - 10%
JEWEL
YaJEM
Correlated uncertaintyShape uncertainty
)c (GeV/T,jetp50 100
| < 0.5jet
η = 0.2 | R TkAnti- c > 5 GeV/ lead,chTp
Data 10 - 30%
JEWEL
YaJEM
Correlated uncertaintyShape uncertainty
ALI−PUB−92182
high-pT hadrons strong suppression : RAA~0.2proxy for Jet (parton) : pT >10 GeV/cfragmentation pattern changed
Jets strong suppression : RAA~0.4Jet shape broadens? where is the lost energy?
6
R
AA
=1
< N
coll
>· #of particles(jets) in PbPb
#of particles(jets) in pp
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
ALICE Jet Quenching Measurements in Pb-PbNuclear modification factor : RAA
if RAA = 1, NO modification of the matter
)c (GeV/Tp
0 2 4 6 8 10 12 14 16 18 20
AAR
0.2
0.4
0.6
0.8
1=2.76 TeVNNsALICE 0-5% Pb-Pb
-π++π-+K+K
pp + Charged
ALI−DER−86149)c (GeV/
T,jetp
0 50 100
AAR
0
0.2
0.4
0.6
0.8
1
1.2 = 2.76 TeVNNsALICE Pb-Pb
Data 0 - 10%
JEWEL
YaJEM
Correlated uncertaintyShape uncertainty
)c (GeV/T,jetp50 100
| < 0.5jet
η = 0.2 | R TkAnti- c > 5 GeV/ lead,chTp
Data 10 - 30%
JEWEL
YaJEM
Correlated uncertaintyShape uncertainty
ALI−PUB−92182
high-pT hadrons strong suppression : RAA~0.2proxy for Jet (parton) : pT >10 GeV/cfragmentation pattern changed
Jets strong suppression : RAA~0.4Jet shape broadens? where is the lost energy?
7
R
AA
=1
< N
coll
>· #of particles(jets) in PbPb
#of particles(jets) in pp
Jet measurement in Pb-Pb collisions is challenging because of very large fluctuating background.
Need a dedicated trigger algorithm to select events containing jets [My first thesis topic]Extract a ‘clean signal’ despite pT smearing and “fake” jets [My second thesis topic]
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Jet Measurement in LHC-ALICE
Calorimeters trigger Jet events &high-pT photon event
L1 trigger implementation
EMCal
DCal
PHOS
ALICE detector focus on Heavy Ion ExperimentLHC Run2 period started from 2015
√s = 13 TeV p-p√sNN = 5.02 TeV Pb-Pb
Charged Particles : |η| < 0.9, 0<φ<2π
EMCal, (DCal : Run2 from 2015-)Pb-Scintillator sampling calorimeter
PHOSlead-tungsten crystal (PWO) based calorimeter
Neutral constituents
Neutral particles : |η| < 0.7
ITS : silicon tracking detectorTPC : gas detector Charged constituents
Full Jet
8
PHOSCharged Jet
Development of the L1-Jet/photon Trigger System by Calorimeters
9
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Search events with energy deposit larger than threshold in certain areaL1 Jet patch ( = 8x8 EMCal modules )
Amplitude calculation by sliding window algorithm
L1-Jet Trigger Algorithm
L1 jet patch = 8x8 modules
jet primitive = 4x4 modules
10
EMCal
η
φ
DCal DCalPHOS
7
3
11
4
0
8
15
12
2
27
35
32
Amplitudes of PHOS jet primitive are scaled complicated geometry of DCal+PHOS side
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
L1-Jet Background Subtraction Method in Pb-Pb
11
Large background energy deposit superimposed to the jet energyevent-by-event correction using the background measured on the opposite side
To avoid jet bias on background estimateTo obtain a reliable event-by-event background
EMCAL
Module BKG_density (EMCAL)
threshold
L1 out
8x8 Module Sum Calculate BKG
Subtract BKG Compare
L1 out
DCAL
Module BKG_density (DCAL)
threshold
8x8 Module Sum Calculate BKG
Subtract BKG Compare
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
PHOS TRUPHOS TRUPHOS TRUPHOS FEE
PHOS TRUPHOS TRU
L1 System of ALICE Calorimeters
PHOS STU
DCAL STU
CTPPHOS TRUPHOS TRU
DCAL TRUDCAL TRUDCAL TRUDCAL TRU
DCAL_L0 DCAL+PHOS_L1_Jet_low
DCAL+PHOS_L1_Jet DCAL_L1_photon_low
DCAL_L1_photon
L0_local fastOR amp
L0_local fastOR amp
subregion data*28TRUs
*14TRUs
DCAL TRUDCAL TRUDCAL TRUEMCAL
TRU
L0_local fastOR amp
*32TRUs
EMCAL STU
background density
EMCAL_L0 EMCAL_L1_Jet_low
EMCAL_L1_Jet EMCAL_L1_photon_low
EMCAL_L1_photon
PHOS_L0 PHOS_L1_photon_low
PHOS_L1_photon
DCAL TRUDCAL TRUDCAL TRUDCAL FEE
DCAL TRUDCAL TRUDCAL TRUEMCAL
FEE
fastOR amp (analog)
fastOR amp (analog)
fastOR amp (analog)
Analog sum inside 1fastOR
digitise fastOR Amp L0 trigger calculation
L1 photon/Jet Trigger calculation
Contribution of Grenoble Group :
Design / Production of the STU (Summary Trigger Unit)
FPGA FW development on STU at period of Run1 (2010-) Olivier. B
MY ACTIVITY :
FPGA FW development on STU for Run2 period (2015-)
Trigger system commissioning
12
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Trigger PerformanceTrigger performance
Data produced by Pb-Pb run in 2015PHOS didn’t join in whole Jet Trigger calculation due to insufficient tune of TRUs
BKG density correlation between EMCal and DCal Reasonable positive correlation
Trigger rejection To define threshold which satisfy the BandWidth restriction of the data taking~1000 for L1-Jet @20GeV threshold
Threshold[GeV]40− 20− 0 20 40 60 80 100
Rej
ectio
n5−10
4−10
3−10
2−10
1−10
1 L1-Jet RejectionEMCalDCal
(EMCAL)[GeV]ρ0 50 100 150 200 250 300
(DCAL)[G
eV]
ρ
0
50
100
150
200
250
300
1
10
210
310
410
510
HIROKI YOKOYAMA
Jet Trigger Rejection
13
BKG density correlation
DCal
BKG
den
sity
EMCal BKG density
Trigger Threshold [GeV]
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Trigger Performance
FEE amp in JP/GP [GeV]0 10 20 30 40 50 60 70 80
effic
ienc
y
0
0.2
0.4
0.6
0.8
1
1.2
HIROKI YOKOYAMA
EMCAL Jet_H
DCAL Jet_H
EMCAL Gamma_H
DCAL Gamma_H
Trigger efficiency clear turn-on at set values of 10 GeV (L1-photon) and 20 GeV (L1_Jet)
All L1-Jet/photon Triggers by Calorimeters Worked Well in 2015 Pb-Pb Runs!
14
Data produced by Pb-Pb run in 2015PHOS didn’t join in whole Jet Trigger calculation due to insufficient tune of TRUs
BKG density correlation between EMCal and DCal Reasonable positive correlation
Trigger rejection To define threshold which satisfy the BandWidth restriction of the data taking~1000 for L1-Jet @20GeV threshold
effici
ency
Energy in Jet/photon patch
Measurement of the Inclusive Jet with √sNN = 5.02 TeV Pb-Pb Collisions
15
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
(1) Jet ReconstructionCorrespondence between detector level, hadron level and parton level.Jet Reconstruction
using observable particlessequentially combine or classify these particles into clusters (Jet candidates)
Some types of algorithms
16
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Analysis Strategies
Charged Particle Selection
Jet Reconstruction
Underlying Event background subtraction
Unfolding
Charged Particle Selection
Neutral Particle Selection
Jet Reconstruction
Unfolding
Inclusive Jet spectra Inclusive Jet spectra
MB Trigger event, Charged Jet MB, Jet Trigger event, Full Jet (Charged+Neutral)
Data : Pb-Pb (√sNN = 5.02 TeV) taken in 2015 Nov. and Dec.
17
Underlying Event background subtraction(2)
(3)
(1)
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Estimated BKG is subtracted from each Jetevent-by-event basis via
background density : ρmedian of kT clusters except largest two:
large background superimposed to the JetsdNch/dη ~ 2000 in most central collisions
[GeV/c]jetT
p0 50 100 150 200 250
T/d
pje
t d
Nev
1/N
5−10
4−10
3−10
2−10
1−10
1
10
210=5.02TeV, cent: 0-10%NNsPbPb
before BKG subtraction
after BKG subtraction
10
210
310
Centrality (%)0 10 20 30 40 50 60 70 80 90
(GeV
/c)
ρ
0
50
100
150
200
250
my result
Difficulties in Heavy Ion Collision
Jet Background Subtraction ← charged jet BKG density vs. centrality
↓ Jet pT spectra before/after BKG subtraction (R=0.4)
my result
18
(2) Underlying Event
0-10 % Central eventsbefore subtractionafter subtraction
~50 GeV BKG in the Jet Area
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Background Fluctuations Detector Effects
(3) UnfoldingMore correction NEEDED!
Detector effect (e.g. smearing) background fluctuation
Unfolding technique
Mm = Gm,d ·Dd Dd = Gd,t · Tt
Measuredjet spectrum
Responsematrix
“unknown” truejet spectrum
Responsematrix
spectrum corrected for BKG fluctuation
Inverted Response matrix gives “TRUE” jet spectrum via
Mm = Gm,d ·Gd,t · Tt = Gm,t · Tt
19
MC simulation
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
SummaryQuark-Gluon Plasma (QGP)
Deconfined state of quarks and gluonsCreated in “Little Bang” ( Heavy Ion Collisions )
Jet quenching Attenuation of the energy of hard scattered partonBest equivalent to quarks/gluons
Development of L1-Jet/photon trigger system by ALICE Calorimeters Background subtraction method was implemented on STU.All work was achieved.
Measurement of Inclusive Jets in HI collisions Need some experimental technique ( jet reconstruction, underlying event subtraction and unfolding )work in progress
Charged Jet Spectra and nuclear modification factor : until Sep. 2016Full Jet Analysis : depend on the EMCal calibration status
20
BACKUP
21
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Relativistic Heavy Ion Collision
Hard parton scatteringhigh momentum parton production
Creation of a Quark-Gluon Plasmathermalisation of strongly interacting partonscollective / thermodynamical properties
Phase transitionfrom parton phase to hadron phaseTc ≈ 155MeV, εc ≈ 0.5GeV/fm3
Hadronic phasefix species, number and momentum of hadrons
Final particles (hadrons, leptons) are detected by particle detectorC.)NONAKA)) ®½Ä°¾�¿KO�)
´±ÁÂ�ÃUC-H��F�thermaliza/on� hydro� hadroniza/on� freezeout�collisions�
22
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Modelling of the Jet Quenching
Radiative energy loss (gluon bremsstrahlung) is dominantRelative strength is controlled by Casimir factor CR
αsCF : q→qgαsCA : g→ggαsTF : g→qqbar
ΔE measurement provides information on matter properties (Jet tomography)
23
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Jet / charged particle spectra in Pb-Pb and p-p collisions
(GeV/c)T
p0 5 10 15 20
-2) (
GeV
/c)
T d
pη
) / (d
chN2) (
dT
pπ
1/(2
evt
1/N
-810
-710
-610
-510
-410
-310
-210
-110
1
10
210
310
410
510
scaled pp reference0-5%70-80%
= 2.76 TeVNNsPb-Pb
ALI−PUB−86029
AWAITI
NG APP
ROVAL
)c (GeV/T,jetp0 50 100
-1 )c (G
eV/
jet
ηdT,
jet
pdN2 d
ev
tN1
coll
N1
-910
-810
-710
-610
= 2.76 TeVNNsALICE
| < 0.5jet
η = 0.2 | R TkAnti-c > 5 GeV/lead,ch
Tp
Correlated uncertaintyShape uncertainty
pp0 - 10% Pb-Pb10 - 30% Pb-Pb
ALI−PUB−92515
24
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
FEE, TRU and STUFEE (Front-End Electronics card)
mainly, calculate analog sum of 2x2 towers amplitudefastOR ( module ) = 2x2 towers area
TRU ( Trigger Region Unit )digitise analog fastOR amplitudeL0 calculation
STU ( Summary Trigger Unit )L1-Jet/Photon calculation
FEE TRU STU
25
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
L1-photon patch 2x2 [modules] (EMCAL, DCAL) 2x2 [2x2 crystals] (PHOS) on-the-fly sliding window algorithm
L1 photon/Jet algorithm in Run20 1 2 3 4 5 6 7
8 9 10 11 12 13 14 15
16 17 18 19 20 21 22 23
24 25 26 27 28 29 30 31
32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47
48 49 50 51 52 53 54 55
56 57 58 59 60 61 62 63
64 65 66 67 68 69 70 71
72 73 74 75 76 77 78 79
80 81 82 83 84 85 86 87
88 89 90 91 92 93 94 95
0 1 2 3 4 5 6 7
8 9 10 11 12 13 14 15
16 17 18 19 20 21 22 23
24 25 26 27 28 29 30 31
32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47
48 49 50 51 52 53 54 55
56 57 58 59 60 61 62 63
64 65 66 67 68 69 70 71
72 73 74 75 76 77 78 79
80 81 82 83 84 85 86 87
88 89 90 91 92 93 94 95
0 1 2 3 4 5 6 7
8 9 10 11 12 13 14 15
16 17 18 19 20 21 22 23
24 25 26 27 28 29 30 31
32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47
48 49 50 51 52 53 54 55
56 57 58 59 60 61 62 63
64 65 66 67 68 69 70 71
72 73 74 75 76 77 78 79
80 81 82 83 84 85 86 87
88 89 90 91 92 93 94 95
0 1 2 3 4 5 6 7
8 9 10 11 12 13 14 15
16 17 18 19 20 21 22 23
24 25 26 27 28 29 30 31
32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47
48 49 50 51 52 53 54 55
56 57 58 59 60 61 62 63
64 65 66 67 68 69 70 71
72 73 74 75 76 77 78 79
80 81 82 83 84 85 86 87
88 89 90 91 92 93 94 95
26
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
L1 photon/Jet algorithm in Run2L1-photon patch
2x2 [modules] (EMCAL, DCAL) 2x2 [2x2 crystals] (PHOS) on-the-fly sliding window algorithm
L1-Jet patch (jet-primitive = 4x4 modules) 2x2 jet-primitives sliding window algorithm
27
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
L1 photon/Jet algorithm in Run2L1-photon patch
2x2 [modules] (EMCAL, DCAL) 2x2 [2x2 crystals] (PHOS) on-the-fly sliding window algorithm
L1-Jet patch (jet-primitive = 4x4 modules) 2x2 jet-primitives sliding window algorithm
L1-Jet background BKG patch = 2x2 jet-primitives no overlap BKG = median of BKG patch amp
Epatch - median{EBKG/ABKG} * Apatch > Threshold => L1 activated
28
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
EMCAL
DCAL
L1-Jet Background Subtraction Method
BKG Jet patch (no overlapped 2x2 jet
primitive)jet primitive calc
BKG density calculation
(median calc)
Module amplitude BKG_density(EMCAL)
BKG subtraction
BKG Jet patch (no overlapped 2x2 jet
primitive)jet primitive calc
BKG density calculation
(median calc)
Module amplitude BKG_density(DCAL)
BKG subtraction
threshold
comparison L1 out
threshold
comparison L1 out
jet primitive
BKG jet patch
BKG jet patch
Jet patch calc (nxn jet primitive)
Jet patch calc (nxn jet primitive)
jet patch
jet patch
PHOS_subregion
jet primitive
jet primitive
jet primitive
29
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Trigger Rejection, Trigger EfficiencyTrigger performance
Threshold[GeV]40− 20− 0 20 40 60 80 100
Rej
ectio
n
5−10
4−10
3−10
2−10
1−10
1 L1-Jet RejectionEMCalDCal
Threshold[GeV]10− 5− 0 5 10 15 20 25 30
Rej
ectio
n
5−10
4−10
3−10
2−10
1−10
1 L1-Photon Rejection
EMCal
DCal
Photon Trigger Jet Trigger
FEE amp in JP/GP [GeV]0 10 20 30 40 50 60 70 80
eff
icie
ncy
0
0.2
0.4
0.6
0.8
1
1.2
HIROKI YOKOYAMA
EMCAL Jet_H
DCAL Jet_H
EMCAL Gamma_H
DCAL Gamma_H
Trigger E�ciency =Number of Jet Patch in Triggered Events
Number of Jet Patch in MB Events
30
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Jet Reconstruction AlgorithmFastJet anti-kT algorithm ( p=-1, p=1 for kT algorithm)
calculate dij and diB by all particles combinationwhen minimum “d” among them is part of dij
merge particle “i” and “j”when minimum “d” among them is part of diB
that cluster defined as jetrepeat until no particle are left
31
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Definition of the Underlying Energy Density1. Output of kT algorithm : the list of the kT clusters ( pi, Ai )
kT algorithm is better than anti-kT algorithm for background estimation because low-pT particles are apt to be the cluster constituents.
2. Remove largest two clusters from the listbecause (2nd) largest cluster is the real Jet candidate.
3. Calculate median of “pTi / Ai ”“median” is better than “average” because the average calculation tend to be affected by seeped real-jet signal
32
Hiroki Yokoyama : 2nd year Thesis Monitoring Seminar, 21/04/2016
Underlying Event @ other centralityat the peripheral collision, background subtraction don’t have impact due to the small background
[GeV/c]jetT
p0 50 100 150 200 250
T/d
pje
t d
Nev
1/N
5−10
4−10
3−10
2−10
1−10
1
10
210=5.02TeV, cent: 0-10%NNsPbPb
before BKG subtraction
after BKG subtraction
[GeV/c]jetT
p0 50 100 150 200 250
T/d
pje
t d
Nev
1/N
5−10
4−10
3−10
2−10
1−10
1
10
210=5.02TeV, cent: 70-90%NNsPbPb
before BKG subtraction
after BKG subtraction
33