Sherpa - Overview and Recent Developments · BSM Capabilities of Sherpa BSM model input via universal UFO format Support for spin 0=1 2 =1 particles Arbitrary Lorentz- and color structures

Sherpa

Overview and Recent Developments

Silvan Kuttimalai

April 20th, 2018

MC4BSM 2018 Durham

Overview

What is Sherpa? [Gleisberg et al.: JHEP 0902 (2009)]

Multi-purpose MC event generator for high-energy collider physics

From hard process to hadron-level events

Focus on perturbative aspects

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

1

Overview

Hard Process

Two tree-level ME generators

Comix [Gleisberg et al.: JHEP 0812 (2008)]

Amegic++ [Krauss et al.: JHEP 0202 (2002)]

Interfaces to loop-ME generators

NLO QCD

NLO EW

Multijet-merging @ LO and NLO

NNLO plugins for few processes

Drell-Yan

Higgs production

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

1

Overview

Dipole Parton Showers

CS shower [Schumann et al.: JHEP 0803 (2008)]

Dire shower [Hoche et al.: Eur.Phys.J. C75 (2015)]

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

1

Overview

Non-Perturbative Aspects

Cluster Fragmentation model

Multi-Parton interaction model

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

1

BSM Physics with Sherpa

BSM Toolchain

Lagrangian

L = LSM + cGΛ2 G

3 + . . .

FeynRules

[Alloul et al., CPC 185 (2014)]

UFO Output

[Degrande et al., CPC 183 (2012)]

Input parameters

Particle spectrum

Vertices

Color structures

Lorentz Structures

Monte Carlo

Sherpa

Herwig

MadGraph

. . .

2

BSM Capabilities of Sherpa

BSM model input via universal UFO format

Support for spin 0/ 12/1 particles

Arbitrary Lorentz- and color structures → EFTs

ME generation with Comix

Spin-correlated decay chains

Model number of max. rel. deviation

processes tested Comix ↔ MadGraph5

Standard Model 60 2.3 · 10−10

Higgs Effective Field Theory 13 4.3 · 10−13

MSSM 401 1.0 · 10−10

Minimal Universal Extra Dimensions 51 2.8 · 10−12

Anomalous Quartic Gauge Couplings 16 5.9 · 10−12

. . .

3

D-6 Effective Gluon Interactions

OG =cGΛ2

fabcGµa,νG

νb,κG

κc,µ

Collider Constraints from tt production

Bound from global top analysis: Λ/√cG > 850 GeV

[Buckley et al., JHEP 04 (2016)]

Idea: use Multijet Events

Two-jet events insensitive to OG

Njet ≥ 4 not considered before

Need to model high-multiplicity signal (5 jets)

NLO not feasible / relevant → use multijet merging

Need very efficient tree-level ME generator (Comix)

[Krauss, SK, Plehn: Phys.Rev. D95 (2017)]4


Describe data well using MEPS@NLO merging

10−1

100

101

102

103

104

Even

tspe

rbi

n

Njet ≥ 4 Λ√cG

= 5 TeV

CMS dataSM

2000 3000 4000 5000 6000 7000ST [GeV]

0.51.01.52.0

Rat

ioto

SM SHERPA

Njet ≥ 4

10−1

100

101

102

103

104

Even

tspe

rbi

n

Njet ≥ 5 Λ√cG

= 5 TeV

CMS dataSM

2000 3000 4000 5000 6000 7000ST [GeV]

0.51.01.52.0

Rat

ioto

SM SHERPA

Njet ≥ 5

5


5.0 5.2 5.4 5.6 5.8 6.0Λ/√

cG [TeV]

10−9

10−8

10−7

10−6

10−5

10−4

10−3

10−2

10−1

100

CL s

Njet ≥ 5 L = 2.2 fb−1

Observed

Expected (±σ)

10−7

10−6

10−5

10−4

10−3

10−2

10−1

dσ

/d

S T[p

b/G

eV]

Njet ≥ 5 Λ√cG

= 5 TeV

SMSM + OG (all terms)

SM + OG (1/Λ2 and 1/Λ4 terms)

SM + OG (1/Λ2 term)

2000 3000 4000 5000 6000 7000ST [GeV]

0.81.62.43.2

Rat

ioto

SM SHERPA

Λ√

cG> 5.2 TeV at 95 % CL

Based on MC for background, potential for improved using data-driven methods

6

Spin-Correlated Decays in the MSSM

u

u

χ02

e+

e−R

χ01

e−

Sher

paM

C

Full MECorrelated decaysUncorrelated decays

0

1

2

3

4

5

6

7

1/σ

dσ

/dm

[10−

3 /GeV

]

0 50 100 150 200 250 300 350

-2-1012

m(u, e+) [GeV]

corr

−fu

ll√

δ2(c

orr)

+δ2

(ful

l)

u

d

χ+1

χ01

W+

µ+

νµ

Sher

paM

C

Full MECorrelated decaysUncorrelated decays

0

1

2

3

4

5

1/σ

dσ

/dm

[10−

3 /GeV

]

0 50 100 150 200 250 300 350 400

-2-1012

m(d, µ+) [GeV]

corr

−fu

ll√

δ2(c

orr)

+δ2

(ful

l)

[Hoche et al.: Eur.Phys.J. C75 (2015)] 7

Recent Developments: NLO EW

NLO EW: Motivation

NNLO QCD → %-level accuracy for many processes α2s ≈ α

Differentially, EW effects can be much larger

Photon radiation off final state leptons

→ Precision Z and W physics

Large corrections at high pT : EW Sudakovs

→ Dark matter backgrounds

Automation in Sherpa

Adapt QCD technology

[Schonherr: Eur.Phys.J. C78 (2018)]

[Lin

der

tet

al.

:E

ur.

Ph

ys.J

.C

77

(20

17

)]

8

NLO EW: Sherpa Results

Sherpa+OpenLoops

V+0,1,2,3 jets [1607.01831/1606.02330/1605.04692]

Z + j/γ + j ratio [1605.04692/1505.05704]

ll + j/lν + j/νν + j/γ + j [1705.04664]

V + H [1607.01831]

llνν [1705.00598]

ttH [1605.04692]

tt+0,1 jet [1803.00950]

Sherpa+GoSam

γγ+0,1,2 jets [1706.09022]

γγγ/γγνl/γγll [1710.11514]

Sherpa+Recola

V jets, llll , ttH [1704.05783]

9

EW Corrections for Top Pairs with Jets

[Gutschow et al.: 1803.00950]

dσNLO EWvirt =[BQCD + VEW +

∫Rsoft dφ1

]dφB

EWvirt Approximation

Captures leading EW Sudakovs

QED real radiation integrated over

Can be combined with QCD PS

Can be used in multijet merging

10

EW Corrections for Top Pairs with Jets

[Gutschow et al.: 1803.00950]EWvirt Approximation

Captures leading EW Sudakovs

QED real radiation integrated over

Can be combined with QCD PS

Can be used in multijet merging

10

Recent Developments:

NLO for Loop Induced Processes

Loop-Induced Processes at NLO

General Treatment

Born and real corrections: Automated one-loop tools

IR-subtraction: standard techniques, e.g. Catani-Seymour

Parton shower matching: standard techniques, e.g. MC@NLO/Powheg

Availability of two-loop virtual amplitudes

gg → γγ [Bern et al.: hep-ph/0109078]

gg → VV → llll [Gehrmann et al.: 1503.04812, Manteuffel et al.: 1503.08835]

gg → HH [Borowka et al.: 1608.04798]

gg → Hj [Jones et al.: 1802.00349]11

Loop-Induced Processes at NLO

Implementation in Sherpa

Use Catani-Seymour for IR subtraction

External 1-loop building blocks (OpenLoops . . . )

Match to parton showers via MC@NLO (CS shower/Dire shower)

[Jones, SK: JHEP 1802 (2018)]

12

Example: Higgs Pair Production

[Heinrich et al.: JHEP 1708 (2017)]

Indications for large uncertainties

13

Higgs Pair Production at LO

[Jones, SK: 1711.03319]

10−11

10−10

10−9

10−8

10−7

10−6

10−5

10−4

10−3

10−2

dσ

/d

pHH⊥

[pb/

GeV

] √s = 14 TeV

Full SM

SHERPA+HHGRID+OPENLOOPS

pp→ HH + jpp→ HH + CS showerpp→ HH + CS shower, µPS =

√s

101 102 103

pHH⊥ [GeV]

0.8

1.6

Rat

io

Sharply falling fixed-order spectrum → PS overshoots tail

14

Higgs Pair Production at NLO

10−9

10−8

10−7

10−6

10−5

10−4

10−3

10−2

dσ

/d

pHH⊥

[pb/

GeV

]

p p → H H√s = 14 TeV

Full SM


Fixed-Order NLOMC@NLO, Dire showerMC@NLO, CS shower

101 102 103

pHH⊥ [GeV]

0.8

1.6

Rat

ioto

FO

0.0

0.1

0.2

0.3

0.4

dσ

/d

pHH⊥

[fb/

GeV

]

p p → H H√

s = 14 TeV

Full SM


NLO+NLL [Ferrera, Pires, 2017]

MC@NLO, Dire showerMC@NLO, CS shower

101 102

pHH⊥ [GeV]

0.7

0.9

1.1

1.3

Rat

ioto

NLO

+NLL

In MC@NLO, PS contributions to tail subtracted to O(αs)

Remainder large, control through judicious choice of shower scale

Match to fixed-order in high p⊥ tail

Good agreement with analytic resummation in low p⊥ region

[Jones, SK: JHEP 1802 (2018)] 15

Recent Developments: NLO PS

Towards NLO Parton Showers

|M|2P

P = + + . . .

So far only LO kernels in parton showers

First steps: Adapt NLO DGLAP for PS evolution

Formally connect PDF evolution to parton shower framework

Genuine 1→ 3 splittings for q → q′

Soft evolution: treat at LO for now

[Hoche et al.: Phys.Rev. D96 (2017)] q q′

16

Towards NLO Parton Showers

[Hoche et al.: JHEP 1710 (2017)]

17

Summary

Sherpa: general-purpose complete event generation framework

BSM physics via universal UFO format

Arbitrary models of spin 0/ 12 /1 particles supported (EFTs)

Recent developments:

NLO electroweak corrections

NLO parton shower

NLO loop-induced processes

18

Performance

0

0

1

1

2

2

3

3

N

100 100

101 101

102 102

103 103

104 104

105 105

106 106

∆t/

µs

per

PSpo

int

gg→ tt + NgAmegicComixOpenLoopsMadGraph standalone cpp

Documents

Sherpa - Overview and Recent Developments · BSM Capabilities of Sherpa BSM model input via universal UFO format Support for spin 0=1 2 =1 particles Arbitrary Lorentz- and color structures