[PPT] - Probing soft and hard radiations with Z (+ jets) Laurent Favart, PowerPoint Presentation

SLIDE 1

REF2017, Madrid

Probing soft and hard radiations with Z (+ jets)

Laurent Favart, Philippe Gras, Anastasia Grebenyuk, Sandeep Kaur, Qun Wang, Fengwangdong Zhang 13 November 2017

SLIDE 2

Introduction

◮ Z (+ jets) precision measurements crucial for deep understanding and modeling of QCD interactions

◮ standard candle at LHC: high cross section; almost background free;

high precision of the full kinematic reconstruction

◮ important for modeling of the production mechanism involved in the

Higgs boson and new physics ◮ Z (+ jets) process can probe different aspects of QCD effects

◮ test latest higher order calculations and MC based event generators ◮ studying multiple gluon emissions and test the models with high

accuracy of soft gluon resummation

2

SLIDE 3

Theoretical prediction for cross section

◮ MADGRAPH5 AMC@NLO + Pythia8 (denoted as LO MG5 aMC)

◮ LO matrix element up to 4 partons ◮ kT -MLM merging ◮ NNPDF3.0 NLO PDF, CUETP8M1 Pythia8 tune

◮ MADGRAPH5 AMC@NLO + Pythia8 (denoted as NLO MG5 aMC)

◮ NLO matrix element up to 2 partons (LO accuracy for 3 partons) ◮ FxFx jet merging ◮ NNPDF3.0 NLO PDF, CUETP8M1 Pythia8 tune

◮ Z+1 jet fixed order NNLO (Phys. Rev. Lett. 115, 062002)

◮ Correction for hadronization and multiple parton interaction

computed with MADGRAPH5 AMC@NLO + Pythia8

◮ CT14 PDF 3

SLIDE 4

Two models with improved soft gluon resummation treatment:

◮ Geneva 1.0-RC2 + Pythia8 (arXiv:1508.01475)

◮ NNLL’+NNLO matched to PS ◮ Use n-jettiness (PRL 105, 092002 (2010)) to separate N-jet and

inclusive (N+1)-jet region, here τ0 and τ1

◮ τ0 (≡ beam-thrust, particles Ei − |pz,i|) dependence resummed at

NNLL’

◮ dσ≥0j at NNLO, dσ≥1j at NLO, dσ≥2j at LO, ◮ PDF4LHC15 NNLO; αs(mZ ) = 0.118 and 0.1135 (for ME and PS) ◮ Specific Pythia8 tune based on CUETP8M1

◮ DYRes 1.0 (JHEP12(2015)047) (shown for pT of the Z boson for Njets ≥ 0)

◮ NNLL+NNLO fixed order calculation ◮ NNPDF 3.1 NNLO and αs(mZ )=0.118 4

SLIDE 5

Transverse momentum of the Z boson for Njets ≥ 0

5 10 15 20 25 30 35 40 45

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL' DYRes (NNLL + NNLO)

MC study

ll → * γ Z/

(Z) [pb/GeV]

T

/dp σ d

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. unc.

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat.

theo. unc.

⊕

= 0.1135

s

α

Stat.

theo. unc.

⊕

(Z) [GeV]

T

p 10

2

10

3

10

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. unc.

NLO MG5 aMC+ Pythia8 LO MG5 aMC+ Pythia8 Geneva (NNLL’+NNLO) + Pythia8 with αs(mZ ) = 0.118 and 0.1135 (dashed line) DYRes NNLL+NNLO The ratio is taken with respect to NLO MG5 aMC Theoretical uncertainties are included for NLO MG5 aMC and Geneva High pT: ◮ predictions with beyond LO ME agrees with each other ◮ LO MG5 aMC is below NLO MG5 aMC

5

SLIDE 6

Transverse momentum of the Z boson for Njets ≥ 0

5 10 15 20 25 30 35 40 45

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL' DYRes (NNLL + NNLO)

MC study

ll → * γ Z/

(Z) [pb/GeV]

T

/dp σ d

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. unc.

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat.

theo. unc.

⊕

= 0.1135

s

α

Stat.

theo. unc.

⊕

(Z) [GeV]

T

p 10

2

10

3

10

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. unc.

Small pT: ◮ both MG5 aMC are based of Pythia8 CUETP8M1 ◮ all the predictions except Geneva are similar

Parameters

CUETP8M1

Geneva tune MultipartonInteractions:pT0Ref 2.4024 2.27 MultipartonInteractions:ecmPow 0.25208 0.25208 MultipartonInteractions:expPow 1.6 1.6 MultipartonInteractions:alphaSvalue 0.130 0.118 (0.1135) ColourReconnection:range 1.8 1.55 SpaceShower:pT0Ref 2.0 1.22 SpaceShower:alphaSvalue 0.1365 0.118 (0.1135) TimeShower:alphaSvalue 0.1365 0.118 (0.1135) BeamRemnants:primordialKThard 1.8 0.32

Primordial kT is reduced in Geneva − → less phase space for parton shower and more from the first principle ◮ Geneva: lower value 0.1135 for αs(mZ ) shows better agreement with NLO MG5 aMC ◮ Geneva: no systematic attempt made to tune the parton shower. Impact of the parton shower tuning need to be understood.

6

SLIDE 7

Jets

◮ By requiring a jet one sensitive to hard gluon radiation in the central region ◮ Possibly to test higher order calculations at high transverse momentum

2 −

10

1 −

10 1 10

2

10

3

10

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL'

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/

[pb]

jets

/dN σ d

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. unc.

jets

N = 0 = 1 = 2 = 3 = 4 = 5 = 6

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat.

theo. unc.

⊕

= 0.1135

s

α

Stat.

theo. unc.

⊕

◮ Jet selection: pT > 30 GeV, |y| < 2.4, separation from the dressed lepton of ∆R > 0.4 ◮ LO MG5 aMC and NLO MG5 aMC show slightly different distribution: more high jet event for NLO MG5 aMC ◮ In Geneva third jet is described by PS. No systematic attempt made to tune the parton shower → disagreements in the central value with the data are expected for Njets≥ 3

7

SLIDE 8

Jet kinematics

Leading jet pT: Subleading jet pT:

2 −

10

1 −

10 1 10

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL' NNLO (1j NNLO)

jetti

N

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/ 1 ≥

jets

ll, N → * γ Z/

) [pb/GeV]

1

(j

T

/dp σ d

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. unc.

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat.

theo. unc.

⊕

= 0.1135

s

α

Stat.

theo. unc.

⊕

) [GeV]

1

(j

T

p 50 100 150 200 250 300 350 400

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat.

theo. unc.

⊕ 2 −

10

1 −

10 1

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL'

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/ 2 ≥

jets

ll, N → * γ Z/

) [pb/GeV]

2

(j

T

/dp σ d

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. unc.

) [GeV]

2

(j

T

p 50 100 150 200 250

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat.

theo. unc.

⊕

= 0.1135

s

α

Stat.

theo. unc.

⊕

◮ LO MG5 aMC has slightly different shape w.r.t NLO MG5 aMC ◮ Leading jet pT: Z+1 jet fixed order NNLO is similar to NLO MG5 aMC within the theory uncertainty; increase theory precision for NNLO calculation ◮ Subleading jet pT: Geneva undershoots NLO MG5 aMC at low pT

8

SLIDE 9

Jet kinematics

Eur. Phys. J. C77 (2017) 361

(ATLAS leading jet pT in Z+jets events )

100 200 300 400 500 600 700 [pb/GeV]

jet T

/dp σ d

7

10

6

10

5

10

4

10

3

10

2

10

1

10 1 10

2

10 ATLAS

1 −

13 TeV, 3.16 fb 1 jet ≥ ) +

−

l

+

l → *( γ Z/ jets, R = 0.4

t

anti-k < 2.5 

jet

y  > 30 GeV,

jet T

p 1 j e t ≥ * + γ Z /

1 −

10 × 2 jets, ≥ * + γ Z/

2 −

10 × 3 jets, ≥ * + γ Z/

3 −

10 × 4 jets, ≥ * + γ Z/ Data NNLO

jetti

1 jet N ≥ Z +

HERPA

S +

AT

H

LACK

B 2.2

HERPA

S 6

Y

P +

LPGEN

A 8 CKKWL

Y

P + MG5_aMC 8 FxFx

Y

P + MG5_aMC

100 200 300 400 500 600 700

Pred./Data 0.5 1 1.5 1 jet ≥ * + γ Z/

(leading jet) [GeV]

jet T

p

100 200 300 400 500 600 700 Pred./Data 0.5 1 1.5 2 jets ≥ * + γ Z/

CMS-PAS-SMP-15-010 (CMS leading jet pT in Z+jets events) ◮ At high pT LO ME model does not describe the measurements → NLO correction is needed ◮ NNLO ME model describes the measurements as good as NLO ME model

9

SLIDE 10

Transverse momentum of the Z boson for Njets ≥ 1

0.5 1 1.5 2 2.5 3 3.5 4

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL'

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/ 1 ≥

jets

ll, N → * γ Z/

(Z) [pb/GeV]

T

/dp σ d

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat. unc.

(Z) [GeV]

T

p 10

2

10

3

10

MG5_aMC Prediction

0.6 0.8 1 1.2 1.4

Stat.

theo. unc.

⊕

= 0.1135

s

α

Stat.

theo. unc.

⊕

At least one jet requirement shift the peak toward the higher value → possibility of studying multiple gluon emissions away from the non-perturbative region High pT: ◮ LO MG5 aMC is below NLO MG5 aMC ◮ Geneva agrees with NLO MG5 aMC Small pT: ◮ Geneva is below NLO MG5 aMC by 20%, the difference is more pronounced than in inclusive scenario ◮ Geneva: use of the lower value 0.1135 for αs(mZ ) improves only the first bin

10

SLIDE 11

Correlation observables

◮ pT balance between the Z boson and the sum of the reconstructed jets: pbal

T

= | pT(Z) +

jets

pT(ji)|, for Njets ≥ 1,2,3 The imbalance is caused by:

◮ hadronic activity outside the jet acceptance (pT > 30 GeV, |y| < 2.4) ◮ gluon radiation in the central region, not clustered in a jet

◮ Jets-Z balance (JZB): JZB = |

jets

pT(ji)| − | pT(Z)| for pT(Z) ≤ 50 GeV and pT(Z) ≥ 50 GeV

◮ the same source of imbalance as for pbal T ◮ it allows the distinction of the two configurations, where

non-accounted hadronic activity is in the Z hemisphere and where it is in the opposite one

11

SLIDE 12

Correlation observable: pbal

T = |

pT(Z) +

jets

pT(ji)|

Njets ≥ 1: Njets ≥ 2: Njets ≥ 3:

2 −

10

1 −

10 1 10

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL'

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/ 1 ≥

jets

ll, N → * γ Z/

[pb/GeV]

T

/dp σ d

MG5_aMC Prediction

0.5 1 1.5

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.5 1 1.5

Stat. unc.

[GeV]

bal T

p 20 40 60 80 100 120 140 160 180 200

MG5_aMC Prediction

0.5 1 1.5

Stat.

theo. unc.

⊕ = 0.1135 s α Stat.

theo. unc.

⊕ 3 −

10

2 −

10

1 −

10 1

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL'

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/ 2 ≥

jets

ll, N → * γ Z/

[pb/GeV]

T

/dp σ d

MG5_aMC Prediction

0.5 1 1.5

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕

[GeV]

bal T

p 20 40 60 80 100 120 140 160 180 200

MG5_aMC Prediction

0.5 1 1.5

Stat. unc.

3 −

10

2 −

10

1 −

10

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 (

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/ 3 ≥

jets

ll, N → * γ Z/

[pb/GeV]

T

/dp σ d

MG5_aMC Prediction

0.5 1 1.5

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕

[GeV]

bal T

p 20 40 60 80 100 120 140 160 180 200

MG5_aMC Prediction

0.5 1 1.5

Stat. unc.

Imbalance (pbal

T

away from zero) from two partons in the final state with one of them

ut of the acceptance - NLO

accuracy for NLO MG5 aMC sample and LO accuracy for

ther samples

Geneva: one jet must come from parton showering LO MG5 aMC shows different shape Smaller difference between LO MG5 aMC and NLO MG5 aMC

12

SLIDE 13

Correlation observable: JZB = |

jets

pT(ji)| − | pT(Z)|

JZB< 0: unaccounted hadronic activity in the Z hemisphere JZB> 0: unaccounted hadronic activity in the opposite hemisphere Njets ≥ 1 is required full phase space: pT(Z) ≤ 50 GeV: pT(Z) ≥ 50 GeV:

3 −

10

2 −

10

1 −

10 1 10

2

10

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL'

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/ 1 ≥

jets

ll, N → * γ Z/

/dJZB [pb/GeV] σ d

MG5_aMC Prediction

0.5 1 1.5

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.5 1 1.5

Stat. unc.

JZB [GeV] 100 − 50 − 50 100

MG5_aMC Prediction

0.5 1 1.5

Stat.

theo. unc.

⊕ = 0.1135 s α Stat.

theo. unc.

⊕ 3 −

10

2 −

10

1 −

10 1 10

2

10

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL'

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/ 1 ≥

jets

ll, N → * γ Z/ 50 GeV ≤ (Z)

T

1, p ≥

jets

ll, N → * γ Z/

/dJZB [pb/GeV] σ d

MG5_aMC Prediction

0.5 1 1.5

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.5 1 1.5

Stat. unc.

JZB [GeV] 40 − 20 − 20 40 60 80 100 120 140

MG5_aMC Prediction

0.5 1 1.5

Stat.

theo. unc.

⊕ = 0.1135 s α Stat.

theo. unc.

⊕ 3 −

10

2 −

10

1 −

10 1 10

2

10

2j NLO + PS) ≤ MG5_aMC + PY8 ( 2j NLO + PS) ≤ MG5_aMC + PY8 ( 4j LO + PS) ≤ MG5_aMC + PY8 ( ) +NNLO

τ

GE + PY8 (NNLL'

MC study

(R = 0.4) Jets

T

anti-k | < 2.4

jet

> 30 GeV, |y

jet T

p ll → * γ Z/ 1 ≥

jets

ll, N → * γ Z/ (Z) > 50 GeV

T

1, p ≥

jets

ll, N → * γ Z/

/dJZB [pb/GeV] σ d

MG5_aMC Prediction

0.5 1 1.5

Stat. theo. ⊕ unc.

s

α ⊕ PDF ⊕ MG5_aMC Prediction

0.5 1 1.5

Stat. unc.

JZB [GeV] 150 − 100 − 50 − 50 100 150

MG5_aMC Prediction

0.5 1 1.5

Stat.

theo. unc.

⊕ = 0.1135 s α Stat.

theo. unc.

⊕

◮ LO MG5 aMC shows different shape w.r.t NLO MG5 aMC ◮ Geneva: similar behavior for three cases

13

SLIDE 14

Correlation observable: JZB = |

jets

pT(ji)| − | pT(Z)|

Effect on the imbalance of the hadronic activity beyond the jet acceptance (pT > 30 GeV, |y| < 2.4): full phase space: pT(Z) ≤ 50 GeV: pT(Z) ≥ 50 GeV:

ZJB [GeV]

200
150
100
50

50 100 150 200

a.u

7

10

6

10

5

10

4

10

3

10

2

10

1

10 1

Default and y cuts

T

No jet p

ZJB [GeV]

50

50 100 150 200

a.u

5

10

4

10

3

10

2

10

1

10 1

ZJB [GeV]

150
100
50

50 100 150

a.u

5

10

4

10

3

10

2

10

1

10 1

Dominant contribution is hadronic activity in the forward region (|y| > 2.4)

14

SLIDE 15

Outlook

DY process allows to study different aspects of QCD ◮ High transverse momentum region:

◮ samples with NLO ME show different behavior than LO ME sample.

Run1 and Run2 measurements show the need of NLO correction to describe proper the high transverse momentum of the Z boson and jets

◮ NNLO ME models are available with significantly reduced theory

uncertainties. Can we access this precision with Run2 measurements?

◮ NNLO ME models do not show large difference with respect to

NLO ME one ◮ Small transverse momentum region:

◮ Run1 and Run2 measurements: precise tune of Pythia8, interfaced

with MG5 aMC, gives very good description of the pT(Z)

◮ NNLO with gluon resummation sample is available where the phase

space for the parton shower with tuned parameters is reduced

◮ Fixed order NNLO+NNLL calculation (DYRes) as good as

models interfaced with Pythia8

◮ NNLL’+NNLO (Geneva) shows different behavior compared to

ther samples

15