1 Studies into Performance and Systematics of ME-PS matching in W+Jets Andrea Messina, CERN Ben Cooper, QMUL Berkeley VB+Jets Workshop 26th March 2008
2 Overview Used Alpgenv2 to study performance and possible systematics of ME-PS matching in W+Jets: –Dependence on PS/UE –Dependence on P T min generation cut –Dependence on CKKW K T factor –Hadron-Parton level relationship All studies done at particle-level and at 1.96 TeV. Plan to extend this study to other approaches and LHC energies - this is a snapshot! Compare predictions to W+Jets measurement (see Andrea’s talk) but also to each other.
3 Details of Study (i) Cross-section definition for predictions identical to our measurement: –E T ele >20 GeV, | ele | 30 GeV, M T W > 20GeV/c 2 –Particle-level predictions clustered using JETCLU cone 0.4. Distributions under study: – (W + n jet) inclusive cross-section jet E T > 25 GeV. –Differential d (W + n jet)/dE T nth jet spectra for jet E T > 20 GeV –Differential d (W + 2 jet)/d R jj jet spectra for jet E T > 15 GeV
4 Details of Study (ii) Constants of our generation: –Alpgen v2.13 matrix element - showering with literally identical generation. –PDF: CTEQ5L –Renorm/Factorisation scale Q = Default generation parameters: –Parton P T min 15 GeV – R jj > 0.4, | jet | < 2.5 –CKKW K T factor = 1.0 “Matching” parameters are always: –Jet E T > (P T min + 5) GeV –R cone = 0.4, | jet | < 2.5 – R(parton-jet) < 0.6
5 Details of Study (iii) Variables: –Parton shower: Pythia(v3.325) or Herwig(v6.510) –For Pythia MSTP(81) = 1 or MSTP(81) = 0 –Parton P T min generation cut –CKKW K T factor –Fully matched or W + np inclusive generation? Keys to different PS/UE generation: –“py” = Pythia (calling PYEVNT, MSTP(81) = 1) –“hw” = Herwig (default tuning - no Jimmy) –“pynoUE” = Pythia with MSTP(81) = 0 UNLESS OTHERWISE STATED IT IS ALWAYS THE ALPGEN MATCHED PREDICTION.
6 Dependence on parton shower/underlying event
7 Herwig & Pythia vs Data PYTHIAHERWIG
8 Herwig & Pythia vs Data PYTHIAHERWIG Substantial difference in predictions at low E T PY/HW
9 Turning UE “off” in Pythia PYTHIA UE Off HERWIG Impact of UE larger for higher njet. PYnoUE/PY
10 Turning UE “off” in Pythia PYTHIA UE OffHERWIG No significant shape change. PYnoUE/PY
11 W+1p Inclusive Generation PYTHIA W+1p inclusive HERWIG W+1p inclusive Pythia/Herwig differences much smaller than for the matched. PY1p/HW1p
12 Herwig & Pythia vs Data: R jj HERWIG PYTHIA PYTHIA W+2p inclusive HERWIG W+2p inclusive
13 Dependence on parton P T min cut
14 Lead d /dE T : Pythia & Herwig HERWIGPYTHIA
15 Herwig Stack Plots: Lead Jet E T P T min 15 P T min 25 P T min 50P T min 8
16 Pythia Stack Plots: Lead Jet E T P T min 15 P T min 25 P T min 50P T min 8
17 Lead Jet E T : Scan of P T min PYTHIA Matched PYTHIA W+1p inclusive
18 Lead Jet E T : Scan of P T min HERWIG Matched HERWIG W+1p inclusive
19 2nd d /dE T : Pythia & Herwig HERWIGPYTHIA
20 Herwig Stack Plots: 2nd Jet E T P T min 25 P T min 50P T min 8 P T min 15
21 Dependence on CKKW vertex K T factor
22 Comparing different K T factors PYTHIAHERWIG See change in njet rates - greater for larger njets
23 Comparing different K T factors PYTHIAHERWIG
24 Comparing different K T factors PYTHIAHERWIG
25 Hadron-level to Parton-level relationship
26 Herwig Inclusive W+1p Matched
27 Pythia UE turned “off” Inclusive W+1p Matched
28 Work-In-Progress Things we have tried but don’t fully understand yet.
29 Herwig
30 Herwig + Jimmy
31 Future plans What is the source of the difference between matched Pythia and Herwig? To what extent are matched predictions independent of generation cuts? Investigate different UE tunes - Pythia v2 Tunes, v3 tunes, Jimmy. Examine further R jj distributions. Look at different approaches: MadEvent, Repeat studies at LHC energies.
32 Backups
33 SMPR vs Alpgen + Pythia
34 SMPR vs Alpgen + Pythia
35 Turning UE “off” in Pythia
36 W + n jet Pythia & Herwig PYTHIA HERWIG
37 Herwig Stack Plots P T min 15 P T min 25 P T min 50
38 Ptmin 25 vs data PYTHIA HERWIG
39 E T > 20 Inclusive
40 Pythia Inclusive W+1p Matched