The Sherpa approach of calculating multijet backgrounds. [Theory - - PowerPoint PPT Presentation

SLIDE 1

The Sherpa approach of calculating multijet backgrounds.

[Theory seminar @ FNAL]

• Jan Winter

a

http://www.sherpa-mc.de/

Aim: improved description of multijet final states Sherpa at a glance CKKW method ... merging tree-level MEs and PSs Survey of application examples ALPGEN vs. Sherpa studies (see EPJC53 (2008) 473) Current developments

a Sherpa authors: T. Gleisberg, S. H¨

• che, F

. Krauss, M. Sch¨

• nherr, F

. Siegert, S. Schumann, J. W.

Jan Winter TH seminar, February 14, 2008 – p.1

SLIDE 2

Challenge – Physics at ...

... future (and present) hadron

and

linear collider experiments. LHC physics: reveal EWSB mechanism, large rates of (B)SM final states LHC: is a QCD machine Multijets huge production phase space Prior to new physics: need to under- stand SM physics/backgrounds V +jets, V V +jets, Q ¯ Q+jets, single t’s VBF and g–g fusion, Higgs production SUSY particles and decay chains

Today’s signals will be tomorrow’s backgrounds.

0.1 1 10 10

• 7

10

• 6

10

• 5

10

• 4

10

• 3

10

• 2

10

• 1

10 10

1

10

2

10

3

10

4

10

5

10

6

10

7

10

8

10

9

10

• 7

10

• 6

10

• 5

10

• 4

10

• 3

10

• 2

10

• 1

10 10

1

10

2

10

3

10

4

10

5

10

6

10

7

10

8

10

9

σjet(ET

jet > √s/4)

LHC Tevatron

σt σHiggs(MH = 500 GeV) σZ σjet(ET

jet > 100 GeV)

σHiggs(MH = 150 GeV) σW σjet(ET

jet > √s/20)

σb σtot

proton - (anti)proton cross sections

σ (nb) √s (TeV)

events/sec for L = 10

33 cm

• 2 s
• 1

CAMPBELL, HUSTON, STIRLING

Jan Winter TH seminar, February 14, 2008 – p.2

SLIDE 3

Challenge – Physics at ...

... future (and present) hadron

and

linear collider experiments. LHC physics: reveal EWSB mechanism, large rates of (B)SM final states LHC: is a QCD machine Multijets huge production phase space Prior to new physics: need to under- stand SM physics/backgrounds V +jets, V V +jets, Q ¯ Q+jets, single t’s VBF and g–g fusion, Higgs production SUSY particles and decay chains

The signal-to-background puzzle. E.g. Higgs in weak boson fusion: Nice rapidity gap. Signal/background ratio depends on central jet veto. Loss of gap structure for higher orders in QCD? Central jet veto to be modified? Backgrounds well modelled? Signal spoiled by UE?

Jan Winter TH seminar, February 14, 2008 – p.2

SLIDE 4

Challenge – Physics at ...

... future (and present) hadron

and

linear collider experiments. LHC physics: reveal EWSB mechanism, large rates of (B)SM final states LHC: is a QCD machine Multijets huge production phase space Prior to new physics: need to under- stand SM physics/backgrounds V +jets, V V +jets, Q ¯ Q+jets, single t’s VBF and g–g fusion, Higgs production SUSY particles and decay chains

The signal-to-background puzzle New-physics discovery signalled by enhanced rate of hard events? Signal: leptons + jets + E

/T .

Is SM backround precisely known? Is it sufficient using PSs only? Jet properties depend on nature of new physics.

Jan Winter TH seminar, February 14, 2008 – p.2

SLIDE 5

Challenge – Physics at ...

... future (and present) hadron

and

linear collider experiments. LHC physics: reveal EWSB mechanism, large rates of (B)SM final states LHC: is a QCD machine Multijets huge production phase space Prior to new physics: need to under- stand SM physics/backgrounds V +jets, V V +jets, Q ¯ Q+jets, single t’s VBF and g–g fusion, Higgs production SUSY particles and decay chains Need for tools that model...!!! Jet production Jet evolution Hadronization

Jan Winter TH seminar, February 14, 2008 – p.2

SLIDE 6

Example: jet mass

BAUR, ORR arXiv:0707.2066

mjet ∝ √αs pjet

T , ...

At higher orders light quark and gluon jets acquire a mass which depends on jet algorithm and

∆R separation.

However, non-perturbative QCD effects may significantly contribute. Before we search for new physics, we want to make sure that detector and reconstruction algorithms operate properly. Jets are always defined according to some algorithm. And different algorithms will give different results.

Jan Winter TH seminar, February 14, 2008 – p.3

SLIDE 7

Monte Carlo event generator Sherpa

• T. Gleisberg, S. Höche, F

. Krauss, A. Schälicke, S. Schumann and J. W., JHEP 0402 056 (2004).

Current version: SHERPA 1.0.11 (released Nov/07).

• ✁

✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✁ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✂ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ✄ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ☎ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✆ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✝ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✞ ✟ ✟ ✟ ✟ ✟ ✟ ✟ ✟ ✟ ✟ ✟ ✟ ✟ ✟ ✟ ✠ ✠ ✠ ✠ ✠ ✠ ✠ ✠ ✠ ✠ ✠ ✠ ✠ ✠ ✠ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ✡ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☛ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✌ ✍ ✍ ✍ ✍ ✍ ✍ ✍ ✍ ✍ ✍ ✍ ✍ ✍ ✍ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✎ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✏ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✑ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✒ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✔ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✕ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✖ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✘ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✙ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✚ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✛ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✜ ✢ ✢ ✢ ✢ ✢ ✢ ✢ ✢ ✢ ✢ ✢ ✢ ✢ ✢ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✣ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✤ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✥ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✦ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ✧ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ★ ✩ ✩ ✩ ✩ ✩ ✩ ✩ ✩ ✩ ✩ ✩ ✩ ✩ ✩ ✪ ✪ ✪ ✪ ✪ ✪ ✪ ✪ ✪ ✪ ✪ ✪ ✪ ✪ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✫ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✬ ✭ ✭ ✭ ✭ ✭ ✭ ✭ ✭ ✭ ✭ ✭ ✭ ✭ ✭ ✭ ✮ ✮ ✮ ✮ ✮ ✮ ✮ ✮ ✮ ✮ ✮ ✮ ✮ ✮ ✮ ✯ ✯ ✯ ✯ ✯ ✯ ✯ ✯ ✯ ✯ ✯ ✯ ✯ ✯ ✯ ✰ ✰ ✰ ✰ ✰ ✰ ✰ ✰ ✰ ✰ ✰ ✰ ✰ ✰ ✱ ✱ ✱ ✱ ✱ ✱ ✱ ✱ ✱ ✱ ✱ ✱ ✱ ✱ ✱ ✲ ✲ ✲ ✲ ✲ ✲ ✲ ✲ ✲ ✲ ✲ ✲ ✲ ✲ ✲ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✳ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✴ ✵ ✵ ✵ ✵ ✵ ✵ ✵ ✵ ✵ ✵ ✵ ✵ ✵ ✵ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✶ ✷ ✷ ✷ ✷ ✷ ✷ ✷ ✷ ✷ ✷ ✷ ✷ ✷ ✸ ✸ ✸ ✸ ✸ ✸ ✸ ✸ ✸ ✸ ✸ ✸ ✸ ✹ ✹ ✹ ✹ ✹ ✹ ✹ ✹ ✹ ✹ ✹ ✹ ✹ ✹ ✹ ✺ ✺ ✺ ✺ ✺ ✺ ✺ ✺ ✺ ✺ ✺ ✺ ✻ ✻ ✻ ✻ ✻ ✻ ✻ ✻ ✻ ✻ ✻ ✻ ✻ ✻ ✻ ✼ ✼ ✼ ✼ ✼ ✼ ✼ ✼ ✼ ✼ ✼ ✼ ✼ ✼ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✽ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✾ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ✿ ❀ ❀ ❀ ❀ ❀ ❀ ❀ ❀ ❀ ❀ ❀ ❀ ❀ ❀ ❀ ❁ ❁ ❁ ❁ ❁ ❁ ❁ ❁ ❁ ❁ ❁ ❁ ❁ ❁ ❁ ❂ ❂ ❂ ❂ ❂ ❂ ❂ ❂ ❂ ❂ ❂ ❂ ❂ ❂ ❃ ❃ ❃ ❃ ❃ ❃ ❃ ❃ ❃ ❃ ❃ ❃ ❃ ❃ ❄ ❄ ❄ ❄ ❄ ❄ ❄ ❄ ❄ ❄ ❄ ❄ ❄ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❅ ❆ ❆ ❆ ❆ ❆ ❆ ❆ ❆ ❆ ❆ ❆ ❆ ❆ ❆ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❇ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❈ ❉ ❉ ❉ ❉ ❉ ❉ ❉ ❉ ❉ ❉ ❉ ❉ ❊ ❊ ❊ ❊ ❊ ❊ ❊ ❊ ❊ ❊ ❊ ❊ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋ ❋

• ❍

❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ❍ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ■ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❏ ❑ ❑ ❑ ❑ ❑ ❑ ❑ ❑ ❑ ❑ ❑ ❑ ❑ ❑ ❑ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▲ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ▼ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ◆ ❖ ❖ ❖ ❖ ❖ ❖ ❖ ❖ ❖ ❖ ❖ ❖ ❖ ❖ ❖ P P P P P P P P P P P P ◗ ◗ ◗ ◗ ◗ ◗ ◗ ◗ ◗ ◗ ◗ ◗ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❘ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❙ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❚ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❯ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❱ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲ ❲

Signal process: AMEGIC++

tree-level ME generator for HP in SM, MSSM, ADD

IS and FS QCD shower: APACIC++

virtuality ordered, Pythia-like showers

ME-PS combination according to CKKW Multiple parton interactions: AMISIC++

à la PRD36:2019,1987; own model under way arXiv:0705.4577

Hadronization: interface to Pythia’s string model;

• wn model under way according to EPJC36:381-395,2004

Hadron decays: interface to Pythia’s decay tool;

• wn comprehensive packages HADRONS++ PHOTONS++ under way

Sherpa is the event generation framework:

• initialization of the different phases
• interplay of the various stages
• steering the event generation
• Jan Winter

TH seminar, February 14, 2008 – p.4

SLIDE 8

Parton showers (PSs)

free colour particle radiates partons perturbatively

annihilation vs. hadronization time: tann ≈ 1/Q : thad ≈ QR2 typical hadron size: R ≈ 0.01 MeV−1 50 GeV quark: tann ≈ 0.02 GeV−1 < < thad ≈ 5 · 103 GeV−1

coll/soft limits ⇒ large logs: AO resummation/LL exponentiation ⇒ tower of logs

O =

• rnαn

S

⇒ O =

• cnαn

S log2n(Q2/Q2 0) + NLL + . . .

factorization – recursive definition in collinear limit dσn+1 = dσn αs(t) 2 π dt t dz Pa→bc(z)

Jan Winter TH seminar, February 14, 2008 – p.5

SLIDE 9

Parton showers (PSs)

• 2

+

• 2

QCD emissions preferably populate collinear and soft phase-space regions propagator factor for q → qg splitting [pq + pg]−2 ≈ [2EqEg(1 − cos θqg)]−2

soft and collinear singularities

cross section factorizes in the collinear limit |Mq¯

qg|2dΦq¯ qg ≈ |Mq¯ q|2dΦq¯ q

αs 2π dtqg tqg Pq→qg(zq) + dt¯

qg

t¯

qg

P¯

q→¯ qg(z¯ q)

• probability for no resolvable emission off quark line between t and t0:

Sudakov form factor (Pq→qg(z) = CF 1+z2

1−z ... spin averaged AP kernel)

∆q(t0, t) = exp

• −

t

• t0

dt′ t′

z+

• z−

dz αs 2π Pq→qg(z)

• ,

z+ = 1 − z− , z− =

• t0/t′

probability for splitting at t1 < t ⇒ dP = ∆q(t1, t) αs

2π 1 t1 Pq→qg(z)dt1dz

Jan Winter TH seminar, February 14, 2008 – p.6

SLIDE 10

Parton showers (PSs)

main features of parton shower approach soft/collinear parton emissions added to initial & final state

[resum LLs]

partons are evolved down to hadronization scale

[ordering in virtuality, angle, pT ]

provides suitable input for universal hadronization models

[scales of O(1 GeV)]

limitations shower seeds are LO QCD processes only lack of high-energetic large-angle emissions semi-classical picture, quantum interferences and correlations only approximate improvements first few hardest emissions according to tree-level MEs

[called ME+PS merging – (L)CKKW, MLM]

use NLO QCD core processes and match to PS

[called MC @ NLO]

Going beyond present shower approximations?

[beyond large NC]

Jan Winter TH seminar, February 14, 2008 – p.7

SLIDE 11

Multiparton tree-level MEs

exact at some fixed order (FO) in the coupling (number of legs) quantum interferences & spin correlations & mass/offshell effects exact phase space filling: correct high energetic/wide angle

configurations

× factorial growth of calculational work

complicated phase-space structures lack of bulk of radiation: multiple soft/coll emissions

1 2 3 n 10 100 1000

Number of Feynman diagrams for qq -> W

+ W

• + n gluons

1 2 3 n 50 100 150 200

Number of parton level processes contributing to pp -> W

+W

• + n jets

u ¯ u W + d W − ¯ νµ µ− e+ νe 1 2 4 3 5 u ¯ u Z0 νe ¯ νe e+ W − ¯ νµ µ− 1 2 3 5 4 u ¯ u Z0 ¯ νµ νµ W + µ− e+ νe 1 3 2 4 5 Graph 1 Graph 2 Graph 3 u ¯ u Z0 e+ e− W − νe ¯ νµ µ− 1 2 3 5 4 u ¯ u e+ e− W − νe ¯ νµ µ− 1 2 3 5 4 u ¯ u Z0 µ− µ+ ¯ νµ W + e+ νe 1 3 2 4 5 Graph 4a Graph 4b Graph 5a u ¯ u µ− µ+ ¯ νµ W + e+ νe 1 3 2 4 5 u ¯ u Z0 W + e+ νe W − ¯ νµ µ− 1 3 5 2 4 u ¯ u W + e+ νe W − ¯ νµ µ− 1 3 5 2 4 Graph 5b Graph 6a Graph 6b

Jan Winter TH seminar, February 14, 2008 – p.8

SLIDE 12

CKKW merging of tree-level MEs and parton showers.

Jan Winter TH seminar, February 14, 2008 – p.9

SLIDE 13

Combine advantages, remove weaknesses. Beware of double counting, preserve universality of hadronization.

matrix element:

• +
• 2

|AR|2 + |BR|2 + 2 Re(ARB∗

R)

parton shower:

• 2

+

• 2

|AR|2 + |BR|2 αS vs. Log

resummed in PS exact ME LO 5jet, but also NLO 4jet

L αn

m NLL exact ME LO 4jet

4 4 4 4 4 5 5 5 5 5 5 5

Jan Winter TH seminar, February 14, 2008 – p.10

SLIDE 14

NLL jet rates

Catani, Dokshitzer, Olsson, Turnock, Webber, Phys. Lett. B269 (1991) 432 Catani, Krauss, Kuhn, Webber, JHEP 0111 (2001) 063

Exclusive e+e− n-jet fractions at cm energy Q and kT resolution Q2

1/Q2

R2(Q1, Q) = [∆q(Q1, Q)]2 , R3(Q1, Q) = 2 [∆q(Q1, Q)]2 Q

Q1

dq Γq(q, Q) ∆g(Q1, q)

∆q(Q1, Q) = exp „ − Z Q

Q1

dq Γq(q, Q) « , Γq(q, Q) = 2CF π αs(q) q „ ln Q q − 3 4 « , Γf (q, Q) = NF 3π αs(q) q ∆g(Q1, Q) = exp „ − Z Q

Q1

dq ˆ Γg(q, Q) + Γf(q) ˜« , Γg(q, Q) = 2CA π αs(q) q „ ln Q q − 11 12 « Sudakov form factors represent probs for q, g to evolve from Q to Q1 with no Q1-resolvable branching.

Jan Winter TH seminar, February 14, 2008 – p.11

SLIDE 15

NLL jet rates

Catani, Dokshitzer, Olsson, Turnock, Webber, Phys. Lett. B269 (1991) 432 Catani, Krauss, Kuhn, Webber, JHEP 0111 (2001) 063

Exclusive e+e− n-jet fractions at cm energy Q and kT resolution Q2

1/Q2

R2(Q1, Q) = [∆q(Q1, Q)]2 , R3(Q1, Q) = 2 [∆q(Q1, Q)]2 Q

Q1

dq Γq(q, Q) ∆g(Q1, q) Improve n-jet distributions above Q1 by replacing ... Γq(q, Q) → |Mq¯

qg|2

Generate distributions below Q1 = Qjet by vetoed parton showering ...

shower started at Q and any emission above Q1 is rejected [to ensure angular ordering] this will e.g. reproduce R2(Q0, Q) = [∆q(Q0, Q)]2: consider one quark line,

∆q(Q1, Q)∆q(Q0, Q)

• 1 +

Q

Q1 dqΓq +

Q

Q1 dqΓq

q

Q1 dq′Γ′ q + . . .

• = ∆q(Q1,Q)∆q(Q0,Q)

exp “ − R Q

Q1 dqΓq

”

, naive showering from Q1 will not reproduce R2: ∆q(Q1, Q)∆q(Q0, Q1) = ∆q(Q0, Q).

Jan Winter TH seminar, February 14, 2008 – p.11

SLIDE 16

CKKW method: phase-space separation

Divide multijet phase space into two regimes by kT jet measure at Qjet. tree-level MEs: jet seed (hard parton) production above Qjet PS: (intra-)jet evolution Qjet < Q < Qcut−off MEs regularized by kT measure requirement Qjet large, unphysical Qjet dependence for fixed multiplicity n, ambiguous phase space

Jan Winter TH seminar, February 14, 2008 – p.12

SLIDE 17

CKKW method: phase-space separation

Divide multijet phase space into two regimes by kT jet measure at Qjet. tree-level MEs: jet seed (hard parton) production above Qjet PS: (intra-)jet evolution Qjet < Q < Qcut−off MEs regularized by kT measure requirement Qjet large, unphysical Qjet dependence for fixed multiplicity n, ambiguous phase space Use k⊥-measure (IRsafe) to define jets.

hadron–hadron collisions:

Qij = min{k2

⊥i, k2 ⊥j} · R2 ij > Qjet

and

QiB = k2

⊥i > Qjet

where R2

ij = 2 [cosh(ηi − ηj) − cos(φi − φj)]

electron–positron collisions:

yij = 2 min{E2

i , E2 j }(1 − cos θij)/S > ycut

Jan Winter TH seminar, February 14, 2008 – p.12

SLIDE 18

CKKW method: reweighting & vetoing

Eliminate/sizeably reduce Qjet dependence at (N)LL. identify pseudo shower history of MEs via backward clustering accordingly reweight MEs by combined αs and Sudakov weight veto PS configurations already included through higher order MEs

Jan Winter TH seminar, February 14, 2008 – p.13

SLIDE 19

CKKW method: reweighting & vetoing

Eliminate/sizeably reduce Qjet dependence at (N)LL. identify pseudo shower history of MEs via backward clustering accordingly reweight MEs by combined αs and Sudakov weight veto PS configurations already included through higher order MEs

∆q(Qjet,Q1) ∆q(Qjet,Q2) ∆q(Qjet,Q1) ∆q(Qjet,Q2) ∆g(Qjet,Q2) Q1 Q2

W=∆g(Qjet,Q2)[∆q(Qjet,Q1)]2 αs(Q2) αs(Qjet) shower history of

• +
• 2

recall ∆q,g : no-emission probability

Jan Winter TH seminar, February 14, 2008 – p.13

SLIDE 20

CKKW method at work

evaluate MEs for 0, 1, . . . , nmax extra partons passing jet criteria at Qjet [regulator, µF,R] select a parton multiplicity with probability Pn = σn/ nmax

i=0 σi

generate parton-level momenta according to the ME reweight ME according to reconstructed pseudo shower history determine parton emission scales Qn, . . . , Q1 using

kT cluster algorithm

calculate corresponding Sudakov weights external partons: ∆q,g(Qjet, Qprod) internal partons: ∆q,g(Qjet, Qprod)/∆q,g(Qjet, Qdec) recalculate αs = αs(Qk) at each cluster tree vertex k start initial-/final-state parton shower for all ME partons at scale where parton was produced and veto shower emissions above Qjet exclusive samples at given resolution scale Qjet inclusive sample with up to nmax “ME” jets by adding them up + highest multiplicity treatment for the nmax MEs

Q1 Q Q1 Q a) b) Q1 Q Q2 Q1 Q Q2 Q Q1 a) b) c)

Jan Winter TH seminar, February 14, 2008 – p.14

SLIDE 21

CKKW – key feature of Sherpa

Method has been implemented within Sherpa in full generality.

• S. Catani, F

. Krauss, R. Kuhn and B. Webber, JHEP 0111 (2001) 063 F . Krauss, JHEP 0208 (2002) 015

Uses built-in ME generator AMEGIC++. Process-independent implementation.

!

Validation W/Z+jets @ Tevatron/LHC

F . Krauss, A. Schälicke, S. Schumann,

• Phys. Rev. D 70 (2004) 114009, D 72 (2005) 054017

W W production @ Tevatron Run II

• T. Gleisberg et al., Phys. Rev. D 72 (2005) 034028

Detailed comparison to MLM merging & Lönnblad-CKKW

Results in EPJC53 (2008) 473

Applications: QCD jets, Zb¯ b + X, VBF, tops, gg → H + X, b-associated Higgs

/ GeV

Z

P 20 40 60 80 100 120 140 160 180 200 10

• 3

10

• 2

10

• 1

1 10

pt Z Z + 0 jet Z + 1 jet Z + 2 jet Z + 3 jet CDF

GeV pb / dP σ d

K = 1.6

Jan Winter TH seminar, February 14, 2008 – p.15

SLIDE 22

CKKW – key feature of Sherpa

Method has been implemented within Sherpa in full generality.

• S. Catani, F

. Krauss, R. Kuhn and B. Webber, JHEP 0111 (2001) 063 F . Krauss, JHEP 0208 (2002) 015

Uses built-in ME generator AMEGIC++. Process-independent implementation.

!

Validation W/Z+jets @ Tevatron/LHC

F . Krauss, A. Schälicke, S. Schumann,

• Phys. Rev. D 70 (2004) 114009, D 72 (2005) 054017

W W production @ Tevatron Run II

• T. Gleisberg et al., Phys. Rev. D 72 (2005) 034028

Detailed comparison to MLM merging & Lönnblad-CKKW

Results in EPJC53 (2008) 473

Applications: QCD jets, Zb¯ b + X, VBF, tops, gg → H + X, b-associated Higgs

/ GeV

Z

P 5 10 15 20 25 30 35 40 45 50 GeV pb / dP σ d 1 10

pt Z Z + 0 jet Z + 1 jet Z + 2 jet Z + 3 jet CDF

K = 1.6

Jan Winter TH seminar, February 14, 2008 – p.15

SLIDE 23

Vary jet separation cut in Sherpa ...

p¯ p → W +W − + X @ Tevatron II: p⊥ of the WW system Qcut ≡ Qjet

SHERPA WW @ Tevatron Run II

=15.0 GeV

cut

Q

WW + X WW + 0jet WW + 1jet WW + 2jets reference

(WW)/GeV]

T

/dlog[p σ ) d σ (1/

• 4

10

• 3

10

• 2

10

• 1

10 1 (WW)/GeV]

T

log[p

• 0.2

0.2

(WW)/GeV]

T

log[p 0.5 1 1.5 2 2.5 SHERPA WW @ Tevatron Run II

=30.0 GeV

cut

Q

WW + X WW + 0jet WW + 1jet WW + 2jets reference

(WW)/GeV]

T

/dlog[p σ ) d σ (1/

• 4

10

• 3

10

• 2

10

• 1

10 1 (WW)/GeV]

T

log[p

• 0.2

0.2

(WW)/GeV]

T

log[p 0.5 1 1.5 2 2.5 SHERPA WW @ Tevatron Run II

=80.0 GeV

cut

Q

WW + X WW + 0jet WW + 1jet WW + 2jets reference

(WW)/GeV]

T

/dlog[p σ ) d σ (1/

• 4

10

• 3

10

• 2

10

• 1

10 1 (WW)/GeV]

T

log[p

• 0.2

0.2

(WW)/GeV]

T

log[p 0.5 1 1.5 2 2.5

Strongly Qcut-dep. subprocesses cooperate total result is decently stable. Residual dependence can be used to tune to a candle process.

Jan Winter TH seminar, February 14, 2008 – p.16

SLIDE 24

Comparison with MCFM’s NLO QCD prediction

J.M. Campbell and R.K. Ellis, Phys. Rev. D 60 (1999) 113006

p¯ p → W +W − + X @ Tevatron II:

• p⊥ of the WW system

0.5 1 1.5 2 2.5 log[pT(W

+W

• ) / GeV]

1e-05 0.0001 0.001 0.01 0.1 1 (1/σ) dσ / d log[pT(W

+W

• ) / GeV]

MCFM NLO (µ=MW) Sherpa 1jet ME level Sherpa 0jet Sherpa 1jet

W

+W

• --> e

+µ

• νeνµ production @ Tevatron Run II

PDF: cteq6l Cuts: pT

lep > 20 GeV, |η lep| < 1.0,

pT

jet > 15 GeV, |η jet| < 2.0,

∆Rll > 0.2, ∆Rlj > 0.4

MCFM @ parton level vs. Sherpa @ shower level: LO ME-level Distribution is described by a delta peak at 0. At NLO the p⊥ of the WW system diverges for soft p⊥s.

Jan Winter TH seminar, February 14, 2008 – p.17

SLIDE 25

Vary scales in Sherpa ... µR and µF

10

• 3

10

• 2

10

• 1

10 10

1

10

2

dσ/dpT [pb/GeV]

Sherpa (default scale choice) Sherpa (all scales x 0.5) Sherpa (all scales x 2)

e

+e

• + X @ LHC

50 100 150 200 250 300 350 400

pT (e

+e

• ) [GeV]
• 0.2

0.2 Qcut=20 GeV nmax=2 50 100 150 200

dσ/dη [pb]

Sherpa (default scale choice) Sherpa (all scales x 0.5) Sherpa (all scales x 2)

e

+e

• + X @ LHC
• 8
• 7
• 6
• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 6 7 8

η (e

+e

• )
• 0.2

0.2 Qcut=20 GeV nmax=2

On the ±20% level only. Much better than pure LO.

Jan Winter TH seminar, February 14, 2008 – p.18

SLIDE 26

Comparison of QCD activity for different MCs

• S. Frixione et al., JHEP 0206 (2002) 029; JHEP 0308 (2003) 007
• T. Sjöstrand et al., CPC 135 (2001) 238

p¯ p → W +W − + X @ Tevatron II: 1: p⊥ of the WW system 2, 3: incl p⊥ of the 1st and 2nd hardest jet

SHERPA WW @ Tevatron Run II

=15.0 GeV

cut

Q

Sherpa 1jet MC@NLO PYTHIA

(WW)/GeV]

T

/dlog[p σ ) d σ (1/

• 4

10

• 3

10

• 2

10

• 1

10 1 (WW)/GeV]

T

log[p

• 0.5

0.5 1

(WW)/GeV]

T

log[p 0.5 1 1.5 2 2.5 SHERPA WW @ Tevatron Run II

=15.0 GeV

cut

Q

Sherpa 1jet MC@NLO PYTHIA

(1st jet) [1/GeV]

T (incl)

/dp σ ) d σ (1/

• 5

10

• 4

10

• 3

10

• 2

10

• 1

10 (1st jet) [GeV]

T (incl)

p

• 0.4
• 0.3
• 0.2
• 0.1

0.1 0.2

(1st jet) [GeV]

T (incl)

p 50 100 150 200 250 300 SHERPA WW @ Tevatron Run II

=15.0 GeV

cut

Q

Sherpa 1jet Sherpa 2jet MC@NLO PYTHIA

(2nd jet) [1/GeV]

T (incl)

/dp σ ) d σ (1/

• 6

10

• 5

10

• 4

10

• 3

10

• 2

10

• 1

10 (2nd jet) [GeV]

T (incl)

p

• 2
• 1

1 2

(2nd jet) [GeV]

T (incl)

p 20 40 60 80 100 120 140 160 180 200

Sherpa vs. MC@NLO and Pythia (shower scale S): comparison of different physics input!

Jan Winter TH seminar, February 14, 2008 – p.19

SLIDE 27

Vary maximal number of tree-level MEs included ...

Sherpa: p¯ p/p → W +W − + X: Scalar sum of lepton & jet transverse momenta.

SHERPA Tevatron II

1/rate= 9.72163 9.976

=15.0 GeV

cut

Q

SHERPA 109 WW+jj 0jet 1jet 2jet SHERPA 109 WW+j 0jet 1jet

[1/GeV]

T

/dH σ ) d σ (1/

• 5

10

• 4

10

• 3

10

• 2

10 [GeV]

T

H SHERPA LHC

1/rate= 0.965059 1.10837

=20.0 GeV

cut

Q

SHERPA 109 WW+jj 0jet 1jet 2jet SHERPA 109 WW+j 0jet 1jet

[1/GeV]

T

/dH σ ) d σ (1/

• 5

10

• 4

10

• 3

10

• 2

10 [GeV]

T

H

• 0.4
• 0.2

0.2 0.4

[GeV]

T

H 50 100 150 200 250 300 350 400 450 500

• 0.4
• 0.2

0.2 0.4

[GeV]

T

H 50 100 150 200 250 300 350 400 450 500

Example of extrapolation to LHC E’s: enhanced QCD radiation phase space. WW + jj vs. WW + j. Slightly different jet pT thresholds for Tevatron/LHC.

Jan Winter TH seminar, February 14, 2008 – p.20

SLIDE 28

Vary maximal number of tree-level MEs included ...

Inclusive jet cross sections at the LHC for Z+jets normalized to the total inclusive cross section:

Monte Carlo

σ≥1jet/σ0 σ≥2jet/σ0 σ≥3jet/σ0 σ=1jet/σ0 σ|y1y2<−2/σ0

CKKW nME = 1

0.304 0.082 0.017 0.222 0.016

CKKW nME = 2

0.340 0.108 0.025 0.231 0.017

CKKW nME = 3

0.348 0.119 0.034 0.229 0.018

Apacic

0.232 0.048 0.007 0.157 0.010

Various Sherpa 1.0.10 predictions are shown; Apacic is Sherpa’s pure shower prediction. Jets are defined according to the Run II kT algorithm and required to have pT,jet > 20 GeV.

Jan Winter TH seminar, February 14, 2008 – p.21

SLIDE 29

Vary maximal number of tree-level MEs included ...

F . Krauss, A. Schälicke, S. Schumann, Phys. Rev. D 72 (2005) 054017

∆Φ (azimuthal) separation of the two leading kT jets in Z/γ∗ + X @ LHC.

12

φ 0.5 1 1.5 2 2.5 3

12

φ /d σ d σ 1/ 0.1 0.2 0.3 0.4 0.5 0.6 0.7 SHERPA

=2

max

n

Z + X Z + 0jet Z + 1jet Z + 2jet =1

max

reference n

12

φ 0.5 1 1.5 2 2.5 3

12

φ /d σ d σ 1/ 0.1 0.2 0.3 0.4 0.5 0.6 0.7 SHERPA

=3

max

n

Z + X Z + 0jet Z + 1jet Z + 2jet Z + 3jet =2

max

reference n

12

φ 0.5 1 1.5 2 2.5 3

12

φ /d σ d σ 1/ 0.1 0.2 0.3 0.4 0.5 0.6 0.7 SHERPA

=4

max

n

Z + X Z + 0jet Z + 1jet Z + 2jet Z + 3jet Z + 4jet =3

max

reference n

nmax = 2 nmax = 3 nmax = 4 Reference is n′

max = nmax − 1 (dashed curve).

shower: uncorrelated emissions, accuracy

• n=2 αn

s ln2n(Q2/Q2 0)

merging: has full ME correlations, yields α2

s

• n=0 αn

s ln2n(Q2/Q2 0)

Jan Winter TH seminar, February 14, 2008 – p.22

SLIDE 30

Sherpa validation against Tevatron data.

Jan Winter TH seminar, February 14, 2008 – p.23

SLIDE 31

Comparison with data: DØ Z+jet production results

DØ collaboration, DØ note 5066-CONF Jet multiplicity, data vs. Pythia (left) and Sherpa (right).

1 2 3 4 5 6 1 10

2

10

3

10

4

10 1 2 3 4 5 6 1 10

2

10

3

10

4

10

data w/stat error data w/stat & sys error Pythia range stat Pythia range stat & sys

D0 RunII Preliminary Jet Multiplicity

• Nr. of Events

1 2 3 4 5 6 0.2 1 2 3 4

Jet Multiplicity Data / PYTHIA

1 2 3 4 5 6 1 10

2

10

3

10

4

10 1 2 3 4 5 6 1 10

2

10

3

10

4

10

data w/stat error data w/stat & sys error Sherpa range stat Sherpa range stat & sys

D0 RunII Preliminary Jet Multiplicity

• Nr. of Events

1 2 3 4 5 6 0.2 1 2 3 4

Jet Multiplicity Data / SHERPA

MC predictions are normalized to total number of events observed in data. Large systematic uncertainties arise from low pT jets ⇒ both predictions are in agreement with data. Pythia tends to underestimate the data.

Jan Winter TH seminar, February 14, 2008 – p.24

SLIDE 32

Comparison with data: DØ Z+jet production results

DØ collaboration, DØ note 5066-CONF Boson transverse momentum (left) and pT of the 2nd jet, data vs. Sherpa.

50 100 150 200 250 1 10

2

10

3

10

4

10 50 100 150 200 250 1 10

2

10

3

10

4

10

data Sherpa range

D0 RunII Preliminary (Z) [GeV]

T

p

• Nr. of Events

50 100 150 200 250 0.2 1 2 3 4

(Z) [GeV]

T

p Data / SHERPA

20 40 60 80 100 120 140 160 180 1 10

2

10

3

10 20 40 60 80 100 120 140 160 180 1 10

2

10

3

10

data w/stat error data w/stat & sys error Sherpa range stat Sherpa range stat & sys

D0 RunII Preliminary jet [GeV]

nd

2

T

p

• Nr. of Events

20 40 60 80 100 120 140 160 180 0.2 1 2 3 4

jet [GeV]

nd

2

T

p Data / SHERPA

MC predictions are normalized to total number of events observed in data. Di-electron system balances the pT of the jet system.

Jan Winter TH seminar, February 14, 2008 – p.25

SLIDE 33

Comparison with data: DØ Z+jet production results

DØ collaboration, DØ note 5066-CONF pT of the 3rd jet, data vs. Pythia (left) and Sherpa (right).

20 40 60 80 100 120 1 10

2

10 20 40 60 80 100 120 1 10

2

10

data w/stat error data w/stat & sys error Pythia range stat Pythia range stat & sys

D0 RunII Preliminary jet [GeV]

rd

3

T

p

• Nr. of Events

20 40 60 80 100 120 0.2 1 2 3 4

jet [GeV]

rd

3

T

p Data / PYTHIA

20 40 60 80 100 120 1 10

2

10 20 40 60 80 100 120 1 10

2

10

data w/stat error data w/stat & sys error Sherpa range stat Sherpa range stat & sys

D0 RunII Preliminary jet [GeV]

rd

3

T

p

• Nr. of Events

20 40 60 80 100 120 0.2 1 2 3 4 5

jet [GeV]

rd

3

T

p Data / SHERPA

MC predictions are normalized to total number of events observed in data.

Jan Winter TH seminar, February 14, 2008 – p.26

SLIDE 34

Comparison with data: DØ Z+jet production results

DØ collaboration, DØ note 5066-CONF Eta and phi difference between the two hardest jets, data vs. Sherpa.

1 2 3 4 5 6 100 200 300 400 1 2 3 4 5 6 100 200 300 400

data w/stat error data w/stat & sys error Sherpa range stat Sherpa range stat & sys

D0 RunII Preliminary (jet,jet) η ∆

• Nr. of Events

1 2 3 4 5 6 0.2 1 2 3 4

(jet,jet) η ∆ Data / SHERPA

0.5 1 1.5 2 2.5 3 50 100 150 200 250 0.5 1 1.5 2 2.5 3 50 100 150 200 250

data w/stat error data w/stat & sys error Sherpa range stat Sherpa range stat & sys

D0 RunII Preliminary (jet,jet) φ ∆

• Nr. of Events

0.5 1 1.5 2 2.5 3

0.2 1 2 3 4

(jet,jet) φ ∆ Data / SHERPA

MC predictions are normalized to total number of events observed in data. Description of angular correlations is ME dominated.

Jan Winter TH seminar, February 14, 2008 – p.27

SLIDE 35

Comparison with data: inclusive dijets @ Tevatron

V.M. Abazov et al., Phys. Rev. Lett. 94 (2005) 221801

Dijet azimuthal decorrelation measured by DØ in Run II. Idea: reconstruct only the 2 leading jets and test soft+hard QCD radiation pattern.

SHERPA SHERPA SHERPA SHERPA SHERPA SHERPA SHERPA SHERPA

Sherpa

/2 π π 3/4 π /2 π π 3/4 π /2 π π 3/4 π /2 π π 3/4 π /2 π π 3/4 π /2 π π 3/4 π /2 π π 3/4 π /2 π π 3/4 π

> 180 GeV (x8000)

max T

p < 180 GeV (x400)

max T

130 < p < 130 GeV (x20)

max T

100 < p < 100 GeV

max T

75 < p

dijet

φ ∆ /d

dijet

σ d

dijet

σ 1/

• 3

10

• 2

10

• 1

10 1 10

2

10

3

10

4

10

5

10

dijet

φ ∆

Jan Winter TH seminar, February 14, 2008 – p.28

SLIDE 36

Comparison between merging approaches.

CERN-PH-TH/2007-066 LU-TP 07-13 KA-TP-06-2007 DCPT/07/62 IPPP/07/31 SLAC-PUB-12604

Comparative study of various algorithms for the merging of parton showers and matrix elements in hadronic collisions ∗

• J. Alwall1, S. H¨
• che2, F. Krauss2, N. Lavesson3, L. L¨
• nnblad3,
• F. Maltoni4, M.L. Mangano5, M. Moretti6, C.G. Papadopoulos7,
• F. Piccinini8, S. Schumann9, M. Treccani6, J. Winter9, M. Worek10,11

1 SLAC, USA; 2 IPPP, Durham, UK; 3 Department of Theoretical Physics, Lund University, Sweden; 4 Centre for Particle Physics and Phenomenology (CP3)

Universit´ e Catholique de Louvain, Belgium;

5 CERN, Geneva, Switzerland; 6 Dipartimento di Fisica and INFN, Ferrara, Italy; 7 Institute of Nuclear Physics, NCSR Demokritos, Athens, Greece; 8 INFN, Pavia, Italy; 9 Institut f¨

ur Theoretische Physik, TU Dresden, Germany;

10 ITP, Karlsruhe University, Karlsruhe, Germany; 11 Institute of Physics, University of Silesia, Katowice, Poland.

Abstract We compare different procedures for combining fixed-order tree-level matrix- element generators with parton showers. We use the case of W-production at the Tevatron and the LHC to compare different implementations of the so-called CKKW and MLM schemes using different matrix-element generators and different parton

• cascades. We find that although similar results are obtained in all cases, there are

important differences.

September 27, 2007

∗Work supported in part by the Marie Curie RTN “MCnet” (contract number MRTN-CT-2006-

035606) and “HEPTOOLS” (contract number MRTN-CT-2006-035505).

EPJC53 (2008) 473 Alpgen Ariadne Helac Madevent Sherpa

Jan Winter TH seminar, February 14, 2008 – p.29

SLIDE 37

Comparison between merging approaches, ...

namely MLM, LL and SHERPA ME-PS-merging, has been accomplished. W + X

dσ/dE⊥1 (pb/GeV) (a) Alpgen Ariadne Helac MadEvent Sherpa 10-3 10-2 10-1 100 101 E⊥1 (GeV)

• 1
• 0.5

0.5 1 50 100 150 200 250 dσ/dE⊥2 (pb/GeV) (b) 10-3 10-2 10-1 100 101 E⊥2 (GeV)

• 1
• 0.5

0.5 1 50 100 150 200 dσ/dE⊥3 (pb/GeV) (c) 10-5 10-4 10-3 10-2 10-1 100 E⊥3 (GeV)

• 1
• 0.5

0.5 1 25 50 75 100 125 150 dσ/dE⊥4 (pb/GeV) (d) 10-5 10-4 10-3 10-2 10-1 100 E⊥4 (GeV)

• 1
• 0.5

0.5 1 25 50 75 100

jet ET spectra at Tevatron Run II

Jan Winter TH seminar, February 14, 2008 – p.30

SLIDE 38

Comparison between merging approaches, ...

namely MLM, LL and SHERPA ME-PS-merging, has been accomplished. W + + X

dσ/dE⊥1 (pb/GeV) (a) Alpgen Ariadne Helac MadEvent Sherpa 10-2 10-1 100 101 102 E⊥1 (GeV)

• 1
• 0.5

0.5 1 50 100 150 200 250 300 350 400 450 500 dσ/dE⊥2 (pb/GeV) (b) 10-2 10-1 100 101 102 E⊥2 (GeV)

• 1
• 0.5

0.5 1 50 100 150 200 250 300 350 400 dσ/dE⊥3 (pb/GeV) (c) 10-3 10-2 10-1 100 101 E⊥3 (GeV)

• 1
• 0.5

0.5 1 50 100 150 200 250 300 dσ/dE⊥4 (pb/GeV) (d) 10-3 10-2 10-1 100 101 E⊥4 (GeV)

• 1
• 0.5

0.5 1 50 100 150 200

jet ET spectra at the LHC similar pattern wrt Tevatron

• nce tuned to

Tevatron data, same extrapolation to LHC can be expected Results in EPJC53 (2008) 473

Jan Winter TH seminar, February 14, 2008 – p.30

SLIDE 39

Comparison between merging approaches, ...

namely MLM, LL and SHERPA ME-PS-merging, has been accomplished. W + + X

(1/σ)dσ/dη1 (a) Alpgen Ariadne Helac MadEvent Sherpa 0.1 0.2 η1

• 0.4
• 0.2

0.2 0.4

• 4
• 3
• 2
• 1

1 2 3 4 (1/σ)dσ/dη2 (b) 0.1 0.2 η2

• 0.4
• 0.2

0.2 0.4

• 4
• 3
• 2
• 1

1 2 3 4 (1/σ)dσ/dη3 (c) 0.1 0.2 η3

• 0.4
• 0.2

0.2 0.4

• 4
• 3
• 2
• 1

1 2 3 4 (1/σ)dσ/dη4 (d) 0.1 0.2 η4

• 0.4
• 0.2

0.2 0.4

• 4
• 3
• 2
• 1

1 2 3 4

jet η spectra at the LHC similar pattern wrt Tevatron

• nce tuned to

Tevatron data, same extrapolation to LHC can be expected Results in EPJC53 (2008) 473

Jan Winter TH seminar, February 14, 2008 – p.30

SLIDE 40

Comparison between merging approaches, ...

namely MLM, LL and SHERPA ME-PS-merging, has been accomplished. W + + X

dσ/dp⊥W (pb/GeV) (a) Alpgen Ariadne Helac MadEvent Sherpa 10-2 10-1 100 101 102 103 p⊥W (GeV)

• 1
• 0.5

0.5 1 50 100 150 200 250 300 350 400 450 500 dσ/dp⊥W (pb/GeV) (b) 25 50 100 250 500 1000 p⊥W (GeV)

• 0.6
• 0.3

0.3 0.6 5 10 15 20 25 30 35 40 45 50 dσ/dη1 (pb) (c) 10 20 30 40 50 60 η1

• 1
• 0.5

0.5 1

• 4
• 3
• 2
• 1

1 2 3 4 (1/σ)dσ/d∆η (d) 0.1 0.2 0.3 0.4 0.5 ∆η

• 1
• 0.5

0.5 1 1 2 3 4 5 6 7 8 9

η1(pT ,1 > 100 GeV) |ηW − η1|(pT ,1 > 40 GeV) pT (W) spectra at the LHC similar pattern wrt Tevatron

• nce tuned to

Tevatron data, same extrapolation to LHC can be expected Results in EPJC53 (2008) 473

Jan Winter TH seminar, February 14, 2008 – p.30

SLIDE 41

Merging tree-level MEs and PSs: the MLM method differs from CKKW mainly in ...

the jet definition used in the MEs; the way of accepting/rejecting jet configurations stemming from the MEs; the details concerning the starting conditions and jet vetoing inside the parton showering. See also studies by Lönnblad, Lavesson (arXiv:0712.2966).

Jan Winter TH seminar, February 14, 2008 – p.31

SLIDE 42

Study of systematics of the merging approaches

Alpgen (+PS by Herwig) (left) vs. Sherpa (right)

Example distributions:

pT of W +, η of 1st jet, ∆R12, differential jet rates

dσ/dp⊥W (pb/GeV) (a) Alpgen ALsc ALpt30 ALpt40 10-2 10-1 100 101 102 103 p⊥W (GeV)

• 0.4
• 0.2

0.2 0.4 100 200 300 400 500 dσ/dp⊥W (pb/GeV) (b) 25 50 100 250 500 1000 p⊥W (GeV)

• 0.4
• 0.2

0.2 0.4 10 20 30 40 50 (1/σ)dσ/dη1 (c) 0.1 0.2 η1

• 0.4
• 0.2

0.2 0.4

• 4
• 3
• 2
• 1

1 2 3 4 (1/σ)dσ/d∆R12 (d) 0.1 0.2 0.3 0.4 0.5 0.6 ∆R12

• 1
• 0.5

0.5 1 1 2 3 4 5 dσ/dp⊥W (pb/GeV) (a) Sherpa SHsc SHkt30 SHkt40 10-2 10-1 100 101 102 103 p⊥W (GeV)

• 0.4
• 0.2

0.2 0.4 100 200 300 400 500 dσ/dp⊥W (pb/GeV) (b) 25 50 100 250 500 1000 p⊥W (GeV)

• 0.4
• 0.2

0.2 0.4 10 20 30 40 50 (1/σ)dσ/dη1 (c) 0.1 0.2 η1

• 0.4
• 0.2

0.2 0.4

• 4
• 3
• 2
• 1

1 2 3 4 (1/σ)dσ/d∆R12 (d) 0.1 0.2 0.3 0.4 0.5 0.6 ∆R12

• 1
• 0.5

0.5 1 1 2 3 4 5 0.2 0.4 0.6 0.8 1 1.2 (1/σ)dσ/dlog10(d1/GeV) (e)

• 0.4
• 0.2

0.2 0.4 0.5 1 1.5 2 2.5 log10(d1/GeV) 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 (1/σ)dσ/dlog10(d2/GeV) (f)

• 0.4
• 0.2

0.2 0.4 0.5 1 1.5 2 2.5 log10(d2/GeV) 0.5 1 1.5 2 2.5 (1/σ)dσ/dlog10(d3/GeV) (g)

• 0.4
• 0.2

0.2 0.4 0.5 1 1.5 2 2.5 log10(d3/GeV) 0.2 0.4 0.6 0.8 1 1.2 (1/σ)dσ/dlog10(d1/GeV) (e)

• 0.4
• 0.2

0.2 0.4 0.5 1 1.5 2 2.5 log10(d1/GeV) 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 (1/σ)dσ/dlog10(d2/GeV) (f)

• 0.4
• 0.2

0.2 0.4 0.5 1 1.5 2 2.5 log10(d2/GeV) 0.5 1 1.5 2 2.5 (1/σ)dσ/dlog10(d3/GeV) (g)

• 0.4
• 0.2

0.2 0.4 0.5 1 1.5 2 2.5 log10(d3/GeV)

Exploring the differences is essential to assess the systematic uncertainties of multijet calculations. Test, validate and tune new MC tools at Tevatron, now!

Jan Winter TH seminar, February 14, 2008 – p.32

SLIDE 43

Inclusive jet cross sections at the LHC

Code

σ[tot] σ[≥ 1 jet] σ[≥ 2 jet] σ[≥ 3 jet] σ[≥ 4 jet]

Alpgen, def 10170 2100 590 171 50 ALpt30 10290 2200 555 155 46 ALpt40 10280 2190 513 136 41 ALpt60 10140 2030 403 93 28 ALscL 10590 2520 790 252 79 ALscH 9870 1810 455 121 33 Sherpa, def 8803 2130 574 151 41 SHkt15 8840 2260 642 175 45 SHkt30 8970 2020 481 120 32 SHkt40 9200 1940 436 98.5 24 SHkt60 9650 1990 431 86.8 19 SHscL 7480 2150 675 205 58 SHscH 10110 2080 489 118 30 SHasL 9095 2366 677 190 53.2 SHasH 8597 1924 486 122 32.1 SHinL 7208 1918 552 156 43.1 SHinH 10347 2310 584 148 39.3

ALpt matching scale variations: ALsc µR variations at vertices

• f clustering:

SHkt merging scale variations: SHsc µR & µF variations: (applied in ME & PS phase) SHas only µR variations: SHin µF variations in initial ME integrations:

Jan Winter TH seminar, February 14, 2008 – p.33

SLIDE 44

News on CKKW: heavy quark production + decays

Narrow width approximation full ME factorizes into production & decay parts AMEGIC++ ... use its decay-chain operation mode

projection onto relevant Feynman diagrams, Breit-Wigner intermediate particle masses

APACIC++ ... enable production + decay showers based on massive splittings

e.g. e+e− → t¯ t FS shower for tops e.g. t → W +b IS shower for top, FS shower for bottom

CKKW ... separate and independent merging of MEs with extra jets & showers in production and any decay CKKW ... reweight and veto by respecting the factorization Schematically, e.g.: p¯ p → t [→ W +bg{1}] ¯ t [→ W −¯ bg{1}] g{1} p¯ p → t [→ W +b] ¯ t [→ W −¯ b] p¯ p → t [→ W +b] ¯ t [→ W −¯ b] g p¯ p → t [→ W +b] ¯ t [→ W −¯ b g] p¯ p → t [→ W +b] ¯ t [→ W −¯ b g] g p¯ p → t [→ W +b g] ¯ t [→ W −¯ b g] g ⇒ “CKKW 1-1-1” ...

Jan Winter TH seminar, February 14, 2008 – p.34

SLIDE 45

News on CKKW: top pair production & decays

Some preliminary LHC results ... pT of t¯ t-system

CKKW 1-1-1 compared to answer given by showering only.

Signal studies

Experimentalists would like to better understand the impact of additional jets (ISR/FSR) and get an estimate on the uncertainty of available MC predictions.

t¯ t+jets as background

Accurate treatment required to specify new-physics searches.

Jan Winter TH seminar, February 14, 2008 – p.35

SLIDE 46

News on CKKW: top pair production & decays

Some preliminary Tevatron II results ... pT of t¯ t-system

CKKW 1-1-1 compared to answer given by showering only.

Signal studies

Experimentalists would like to better understand the impact of additional jets (ISR/FSR) and get an estimate on the uncertainty of available MC predictions.

t¯ t+jets as background

Accurate treatment required to specify new-physics searches.

More CKKW studies.

Jan Winter TH seminar, February 14, 2008 – p.36

SLIDE 47

News on CKKW: top pair production & decays

Sherpa: pp → t¯ t → e+νejjjj: pT of 1st jet & trijet mass of combination 134.

SHERPA SHERPA SHERPA

1st jet

T

TTbar production and decay : p LHC

135.384304313 9.82261167 9.82261 Sherpa shower

1st jet

T

TTbar production and decay : p LHC

157.723384692 10.01501449 10.015 CKKW 1-0-0

1st jet

T

TTbar production and decay : p LHC

205.276595628 6.322813475 6.32281 CKKW W+4jets

[pb/GeV]

T,1

/dp σ d

• 4

10

• 3

10

• 2

10

• 1

10

[GeV]

T,1

p (MC-ref) / ref

• 0.5

0.5

[GeV]

T,1

p

50 100 150 200 250 300 350 400 450 500

SHERPA SHERPA SHERPA

TTbar production and decay : trijet mass jets 134 LHC

278.870386794 9.82261505504 9.82261 Sherpa shower

TTbar production and decay : trijet mass jets 134 LHC

295.023985481 10.0150142827 10.015 CKKW 1-0-0

TTbar production and decay : trijet mass jets 134 LHC

335.190093516 6.32280722846 6.32281 CKKW W+4jets

134

/dm σ d

0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1

134

m (MC-ref) / ref

• 0.5

0.5

134

m

100 150 200 250 300 350 400

require at least 4 jets (pT > 40GeV, |η| < 2.5, D = 0.4); lepton cuts (pT > 20GeV, |η| < 2.5); missing energy (E /T > 20GeV) PRELIMINARY RESULTS !!

Jan Winter TH seminar, February 14, 2008 – p.37

SLIDE 48

CKKW at work – Summary

Improved (leading-order) description of hard multijet configurations together with jet fragmentation ! Way of consistently incorporating QCD corrections provided by real-emission MEs. Avoids most serious problems of double counting and missing phase-space regions. CKKW is implemented for SM processes in Sherpa ⇒ tool for jet physics. Thanks to the built-in tree-level ME generator AMEGIC++, real-emission MEs are easily provided. Shapes are in fairly good agreement with NLO predictions; rates, of course, are not NLO. Evidence that constant K-factors may be sufficient. Fairly process independent implementation. Full simulation of hadron-level events valuable tool for experimentalists.

Jan Winter TH seminar, February 14, 2008 – p.38