Interjet Energy Flow in PHP

1
Interjet Energy Flow in PHP
Patrick Ryan University of Wisconsin Claire
Gwenlan Oxford University June 27, 2005

• Monday Meeting
• http//www-zeus.desy.de/pryan/rap_gap

2
Rapidity Gap Events

• Use pQCD to study diffraction
• Hard Diffractive PHP
• Hard High ET Jets (ET gt 5 GeV)
• Diffractive Gap between jets
• Photoproduction Q2 0
• Rapidity Gap Topology
• Distance between jet centers Dh
• ETGap Total ET between leading and trailing
  jets
• Gap Event ETGap lt ETCut
• Gap indicates color singlet exchange

q
t
2p
g Remnant
f
Leading
Jet
Dijet Events with large Rapidity separation
and ETGap lt ETCut
Gap
ET
Trailing
Jet
p Remnant
0
h
-2.4
2.4
All Dijet Events with large Rapidity separation
3
Simulation of gp EventsZEUS - AMADEUS

• PYTHIA 6.1 and HERWIG 6.1 MC
• Direct and Resolved MC generated separately
• Resolved MC includes Multi Parton Interactions
• Dir and Res combined by fitting xg distributions
  to data
• Color Singlet Exchange MC
• HERWIG BFKL
• Uses BFKL Pomeron as exchange object in Rapidity
  Gap events
• PYTHIA High-t g
• Purpose is simply to match the data
• Note Rapidity Gap not due to photon exchange

4
Event Selection and xgOBS Fitting

• ZEUS 96-97 Data
• Luminosity 38 pb-1
• Offline Cleaning Cuts
• zvtx lt 40 cm
• No Sinistra95 e with
• Pe gt 0.9, Ee gt 5 GeV, ye lt 0.85
• 0.2 lt yjb lt 0.85
• Dijet Selection
• ET1,2 gt 5.1, 4.25 GeV
• h1,2 lt 2.4
• ½h1 h2 lt 0.75
• (Spx)2 (Spy)2 / SET lt 2 GeV1/2
• 2.5 lt h1 - h2 lt 4.0 ? Gap Definition
• 4 Gap Samples
• ETCUT 0.5, 1.0 1.5, 2.0 GeV (Hadron)
• ETCUT 0.6, 1.2 1.8, 2.4 GeV (Detector)
• Different Gap ET
• HPP Trigger
• FLT Slot 42

HERWIG xgOBS Fit to Data
Direct Resolved
Direct
PYTHIA 30 Direct 70 Resolved HERWIG 44
Direct 56 Resolved (Using Tuned HERWIG/PYTHIA
- see later slides)
Mixing used to correct data to had level
5
Gap ET Cross Section Default ZEUS
PYTHIA HERWIG
HERWIG
PYTHIA

• Default MC
• Used to unfold data
• Plotted vs. Data
• MC does not describe data at large Gap ET (region
  with no CS)
• Need good agreement at High Gap ET to establish
  depletion at Low Gap ET

6
Large Systematic Differences Default PYTHIA
HERWIG
Data Corrected with PYTHIA HERWIG

• Large Sys Differences
• Large Systematic Errors
• Tuning Procedure
• Match unfolded data and HZTOOL prediction in
  Highest 3 Gap ET bins
• Region without CS contribution
• Generate AMADEUS using tuned parameters

7
PYTHIA Tuning

• Default ZEUS PYTHIA 6.1
• Proton PDF GRV94, LO (Set 5)
• Photon PDF SaS2D (Set 3 of SaSph)
• pTMin 1 2.0
• pTMin 2 1.5
• Modified (Tuned) PYTHIA 6.1
• Based on JetWeb parameters
• Proton PDF CTEQ 5L (Set 46)
• Photon PDF SaS2D (Set 3 of SaSph)
• pTMin 1 1.9
• pTMin 2 1.7

pTMin 1 pT of Hardest interaction pTMin 2 pT of
all secondary interactions
8
HERWIG Tuning

• Default ZEUS HERWIG 6.1
• Proton PDF GRV94 LO (Set 5)
• Photon PDF WHIT-G 2
• Factor to reduce proton radius 1.0
• Probability of Soft Underlying Event 1.0
• PTMIN1 1.8 GeV
• Modified (Tuned) HERWIG 6.1
• Based on JetWeb parameters
• Proton PDF CTEQ 5L (Set 46 of CTEQ)
• Photon PDF SaS2D (Set 3 of SaSph)
• Factor to reduce proton radius 3.0
• Probability of Soft Underlying Event 0.03
• PTMIN1 2.7 GeV

9
Kinematic Variables - HERWIG
Default HERWIG
Tuned HERWIG

• Tuned HERWIG gives better description of Data
  than default HERWIG

10
Kinematic Variables PYTHIA
Default PYTHIA
Tuned PYTHIA

• Tuned PYTHIA gives comparable description of Data
  
• Now have two MCs that describe data well

11
Gap ET Cross SectionTuned PYTHIA and HERWIG
Unfolded without CS
Unfolded with CS

• Reduced systematic difference between HERWIG
  PYTHIA
• Large Gap ET well described
• Unfolding with CS changes cross section in low
  Gap ET bins 10
• Color Singlet Contributions
• PYTHIA 3.1 HERWIG 3.8

12
Comparison Between P.R C.GET Gap

PYTHIA Gap ET
HERWIG Gap ET

• Data unfolded with CS

• Excellent agreement between analyses

13
Delta Eta Cross Section and Gap Fraction

• MC CS describes data in all regions

14
Comparison Between P.R C.GDelta Eta Inclusive
Cross Section

Delta Eta PYTHIA
Delta Eta HERWIG

• Data unfolded with CS

• Excellent agreement between analyses

15
Comparison Between P.R C.G Delta Eta Gap Cross
Section
Gap Cross Section HERWIG
Gap Cross Section PYTHIA
ETCut 0.5
ETCut 1.0
ETCut 1.0
ETCut 0.5
ETCut 2.0
ETCut 1.5
ETCut 2.0
ETCut 1.5

• Data unfolded with CS

• Good agreement between analyses (not all MC stats
  used)

16
Gap ET Cross Section Unfolded with CS Modified
Binning

• Original ETGap Binning
• Had Det 0.5, 1.5, 3.5, 7.0, 12.0
• ETCut in Dh Cross Section
• Had 0.5, 1.0, 1.5, 2.0
• Det 0.6, 1.2, 1.8, 2.4
• Chosen for max purity efficiency
• Inconsistency between ETGap and Dh cross section
  plots
• Divided Gap ET by 1.2
• Percent CS for modified bins
• PYT 3.10 HER 3.38
• Less systematic difference in lower ETGap bins

17
Comparison Between Bin MethodsET Gap

PYTHIA Gap ET
HERWIG Gap ET

• Change 5 in ETGap cross sections

18
Comparison Between Bin MethodsGap Cross Sections

Gap Cross Section PYTHIA
Gap Cross Section HERWIG
ETCut 0.5
ETCut 0.5
ETCut 1.0
ETCut 1.0
ETCut 1.5
ETCut 2.0
ETCut 2.0
ETCut 1.5

• Very small change in Dh Cross Sections

19
Gap ET Cross Section Request for
Preliminary
Avg Data Did not renormalize
MC
Avg Data Refit MC to Avg Data

• Average of Data unfolded with PYTHIA and HERWIG
• Same data points in both plots, MC curves change

20
Delta Eta Cross Section Request for
Preliminary
Avg Data Refit MC to Avg Data

Avg Data Did not
renormalize MC

• Average of Data unfolded with PYTHIA and HERWIG
• Same data points in both plots, MC curves change

21
Gap FractionRequest for Preliminary
Avg Data Refit MC to Avg Data

Avg Data Did not
renormalize MC

• Average of Data unfolded with PYTHIA and HERWIG
• Same data points in both plots, MC curves change

22
Interjet Energy Flow Summary

• Conclusions
• Tuned HERWIG PYTHIA both describe data well
• High Gap ET well described
• Reduced systematic difference between data
  unfolded with HERWIG and PYTHIA
• Gap ET Dh Cross Section well described
• Evidence of 3-4Color Singlet Exchange
  contribution
• Good agreement between P.R. and C.G. analyses
• Modification of ETGap bins necessary for
  consistency
• Two methods for displaying MC curves - must
  choose one
• Normalize to Data points unfolded separately with
  PYT HER
• Normalize to Average of Data unfolded with PYT
  HER
• Plans
• Attempt to make results preliminary for EPS
• Write paper

23
Old vs. New Results
Preliminary ICHEP 2002
New Results (P.R. and C.G.)

• New Results Better description of data at large
  Dh
• Improves confidence in CS extraction