High pT Multi-jet Calibration for 2011 Data

Introduction

This page is intended for the preparation of 2012 Moriond CONF note of multi-jet balance calibration for high pT jet energy scale (JES). The main twiki page of multi-jet calibration can be found here.

The main aim of this 2012 multi-jet balance calibration CONF note is

• to validate MC-based JES calibration at high pT using in-situ multi-jet balance
• to derive in-situ JES calibration on top of MC corrections using gamma+jets and Z+jets calibration at low pT
• to provide the combination of in-situ JES systematic uncertainties

CONF Note Draft

Not ready yet

Datasets

The main datasets are NTUP_JETMET D3PDs produced from Rel. 17 reprocessed data and MC11 simulation. The full D3PDs were slimmed down into the subset (branch list : Data, MC).

Main datasets :

• Data : data11_7TeV.*.physics_JetTauEtmiss.merge.NTUP_JETMET.*p766/
• MC11b Pythia Jx : mc11_7TeV.1050%.J%_pythia_jetjet.merge.NTUP_JETMET.e815_s1273_s1274_r2923_r2900_p832/
• MC11c Alpgen+Jimmy multi-jets : mc11_7TeV.113%.AlpgenJimmy%jetsNp%_J%.merge.NTUP_JETMET.e%_s1310_s1300_r3043_r2993_p839/

Datasets for systematic uncertainty studies:

• MC11b Herwig++ : mc11_7TeV.1132%.HerwigppJetsJ%.merge.NTUP_JETMET.e873_s1349_s1300_r2923_r2900_p832/
• MC11c Pythia Perugia2011 : mc11_7TeV.126%.J%_pythia_jetjet_Perugia2011.merge.NTUP_JETMET.e877_s1349_s1300_r3060_r2993_p839/

Good Runs List

• JetEtMiss standard GRL : data11_7TeV.periodAllYear_DetStatus-v36-pro10_CoolRunQuery-00-04-08_JetEtMiss.xml
• Total integrated luminosity of period D-M data is 4.77 +/- 0.16 fb-1 (3.4% uncertainty).

Analysis

Trigger and baseline event selections, jet quality cuts and multi-jet event selections are described here.

Run Period and Trigger

• Use 2011 data in periods D-M
• EF_j55_a4tc_EFFS trigger at 210 < pT(recoil) <= 260 GeV
• EF_j100_a4tc_EFFS trigger at 260 < pT(recoil) <= 400 GeV
• EF_j135_a4tc_EFFS trigger at 400 < pT(recoil) <= 500 GeV
• EF_j240_a4tc_EFFS trigger at pT(recoil) > 500 GeV

Baseline Event Selection

• Require at least one primary vertex with tracks.
• Reject event if either a Loose" bad or a ugly jet with pT(calibrated) > 20 GeV is present (see here for bad/ugly definition). Here the "calibrated" pT means the pT at EM+JES or LCW+JES scale.
• Reject event if a good jet falls into the vicinity of LAr hole in period E-H for data and run 183003 for MC by following this procedure (jet pT threshold = 40 GeV).
• Reject event if it contains either a lepton (electron or muon) or a photon. The lepton and photon identification cuts are as follows:
  • Electron
    • el_cl_pt > 15 GeV
    • |el_cl_eta| < 2.47
    • Remove crack : 1.37 < |el_cl_eta| < 1.52
    • el_author == 1 or el_author == 3
    • Medium quality (el_medium == 1)
    • Isolation : el_Etcone20/el_cl_pt < 0.1
    • Pass object quality flag ((el_OQ & 1446) == 0)
  • Muon
    • mu_staco_pt > 15 GeV
    • |mu_staco_eta| < 2.5
    • mu_staco_isCombinedMuon
    • mu_staco_ptcone20/mu_staco_pt < 0.1
  • Photon
    • ph_cl_pt > 15 GeV
    • |ph_cl_eta| < 2.37
    • Remove crack : 1.37 < |ph_cl_eta| < 1.52
    • ph_author == 4
    • Tight quality (ph_tight == 1)
    • Isolation : ph_Etcone20 < 10 GeV + 0.02 * ph_cl_pt
    • Pass object quality flag ((ph_OQ & 34214) == 0)

Jet Collection and Calibration

The following four jets collections are under study:

• Anti-k_T, R = 0.4, topo-cluster, EM+JES scale
• Anti-k_T, R = 0.6, topo-cluster, EM+JES scale
• Anti-k_T, R = 0.4, topo-cluster, LCW+JES scale
• Anti-k_T, R = 0.6, topo-cluster, LCW+JES scale

The corresponding D3PD variables are jet_AntiKt4LCTopo_*, jet_AntiKt6LCTopo_*, jet_AntiKt4TopoEM_* and jet_AntiKt6TopoEM_*.

Jet Selection

• Apply Offset, EtaJES, dijet eta-intercalibration using ApplyJetCalibrationTool (using Rel17_JES_EtaInterCalibration_Feb9_2012.config in ApplyJetCalibration-00-00-09 as of 21/2/2012).
• pT > 25 GeV and |eta_EM|< 2.8(1.2) for sub-leading(leading) jets where the pT cut is on ApplyOffsetEtaJES corrected pT and the rapidity cut is applied at the EM-scale.
• Select only jets which are neither "Medium" Bad nor Ugly jets (see here for the definitions of Bad/Ugly jet.)
• Reject jets with E(calibrated) < 0 GeV.
• Sort jets in descending pT

Multi-jet Event Selection

• Require event to have at least 3 good jets.
• radian, where is the azimuthal opening angle between the highest pT jet and the recoil system
• radian, where is the azimuthal opening angle of a sub-leading jet that is closest to the highest pT jet in , measured with respect to the highest pT jet.
• to ensure that the leading and sub-leading jets are at different pT scale.
• GeV.

Calibration Procedures

Two calibration procedures, which are complement to each other, are studied for this CONF note. The decision is made not to use the method with iteration process. Below is just for the record.

Validating high pT JES without iteration:
This approach is to validate JES at high pT at once without iterations. In this method, either an asymmetric cut or an absolute upper cut is applied to recoil jet pT. We study four different cuts to evaluate systematic uncertainty on the highest pT jet energy scale;

1. GeV for all sub-leading jets
2. 
3. 
4. 

where pT is always at the EM+JES scale. We decided to use the cut as it has the smallest bias to the leading and recoil jet pTs and most efficiently suppresses events with dijet-like event topology.

Validating high pT JES with iteration:
This method includes iterative process to reach a higher pT. Neither an asymmetric nor an absolute upper pT cuts (except the initial cut used in a first iteration) are applied in this method. This calibration has been tested with 3 pb-1 of data and looks working well (presentation in Pisa workshop).

1. Select events with pT < 200 GeV for sub-leading jets
2. Make a 2D histogram of vs. , and the recoil pT axis is binned with a bin width of 100 GeV from 200 GeV
3. Calculate the average of and in each bin of recoil pT up to the recoil bin with Nevents > 10
4. Define the upper bound of this iteration in as the average value of in the highest recoil pT bin
5. Make a TGraph using these points:
   • For each jet with pT < upper bound at the EM scale, read off a correction factor for a given using a spline interpolation and Eval() function
   • Use this factor to correct the jet 4-momentum
6. In the next iteration the sub-leading jet cut is raised to pT < upper bound, and the process is repeated

Plots in the Note

Useful information

Meeting Presentations

• Material for Jet calibration meeting
• Jet calibration meeting, Jan. 10, 2012
• Jet calibration meeting, Nov. 29, 2011
• JetEtMiss meeting, Nov. 9th, 2011
• SLAC workshop, Sep. 20th, 2011

---

Major updates:
-- KojiTerashi - 21-Feb-2012

%RESPONSIBLE% KojiTerashi
%REVIEW% Never reviewed