inclusive jet cross section jet energy corrections anwar ahmad bhatti doe meeting december 2, 2004

20
Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Post on 22-Dec-2015

216 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Inclusive Jet Cross Section Jet Energy Corrections

Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Page 2: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

QCD and Jet Physics

All production processes are “QCD related”

Optimal understanding is basic for all analyses:

- Main parameters (ex: gluon PDFs in high x)

- Non perturbative regime (ex: underlying event studies)

- Studies of specific processes where QCD is important

Probe higher energy scales:

- Precise test of perturbative QCD at NLO

- Look for deviations due to new physics

Run I studies were at 10-20%. Aim to improve them to 5%

a) Requires better understanding of relation between experimental measurements and theory (NLO QCD) predictions, kinematic variables clustering, etc.

b) Non-perturbative corrections, underlying events, hadronization

c) Jet energy scale, but improvements limited by detector, jet fragmentation

Page 3: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Inclusive Jet Cross Section

In Run I, inclusive jet cross section had an excess at high jet Et.

The gluon distribution at high x, driven by D0 forward jet data, has indeed increased in recent PDF fits (CTEQ6, CTEQ6.1) leading to better agreement with the CDF inclusive jet cross section measurements.In Run II, the increase in √s and the large integrated luminosity will increase the kinematic range to 600 GeV.The RunII inclusive jet cross section are consistent with the NLO QCD prediction CTEQ 6.1. The band represents 5% uncertainty in jet energy scale.

Page 4: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Ratio of Data over Theory

The data are in good agreement with the NLO CTEQ predictions within the theoretical and experimental uncertainty bands.There is some indications of the data being below the theory at low ET and somewhat above at higher ET, but overall the agreement is good!

Run I with CTEQ6.1CDF Run II

Page 5: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

MidPoint Cone Clustering Algorithm Add seed at middle of two clusters to remove infra red/collinear singularitiesLess detector dependence at clustering stage.Different/better kinematic variables (4-vector and rapidity, instead of Et and η )New method to correct for energy scale and smearing

Pt of the Jet (GeV)

Data Corrected to hadrons Data Corrected to partons

Fmerge=0.5

Page 6: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Kt Clustering (Parton shower motivated)

The comparison is not fair as data is not corrected for underlying event .

Same trend in PYTHIA and HERWIG

Page 7: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Current Status of Inclusive Jet 350 pb-1 data Group decision to “bless” once the results are final i.e. energy scale

uncertainty is reduced below or same as Run I (~3%). Calorimeter calibration/ Run II corrections in good shape but need

some more work. Working on a different method to correct the measured spectrum.

Compare and study biases. Better understanding of new clustering algorithm. Still evolving as

data has configurations where current MidPoint clustering fails. The papers ready in next 4-6 months.

Page 8: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Jet Energy Scale and Top Mass

Systematic Uncertainties

Mtop(GeV/c2)

Jet Energy Scale 5.3

Transfer function 2.0

MC Modeling(jet,UE) 0.5

ISR 0.5

FSR 0.5

PDF 2.0

Generator 0.6

Spin correlation 0.4

NLO effect 0.4

Bkg fraction (±5%) 0.5

Bkg Modeling 0.5

Total 6.2

2top GeV/cm 2.68.177 5.4

0.5 Dynamical Likelihood Method b-tagging, lepton+4 jets, 162 pb-1 22 events 4 background

Page 9: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Generic Jet Corrections at CDF Jet Energy and Resolution Group (Anwar, Florencia Canelli, Tommaso Dorigo) Charge includes a) generic, b-jet corrections b ) improving di-jet mass resolution geared for dijet/b\bar b resonances. The jet energy scale is determined by convoluting the individual particle response

with particle spectrum. It requires good understanding of a) single particle response, measured in data (test beam) and used by simulation b) particle spectrum in the jets. Use photon-jet and Z-jet as a cross check. Steps in the corrections 1. η-dependent corrections, scale all jets in the event to equivalent jets in central

calorimeter. (Ken Hatakeyama will talk about it.) 2. Subtract energy from additional p\bar p interactions 3. Correct calorimeter-level jets energy to hadron-level jets for central calorimeter

non-linearity. 4. Subtract “underlying event” 5. Add energy outside the clustering cone to recover parton energy.

Page 10: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Calorimeter Response to Jets

•Calorimeter Response to jet is convolution of momentum of hadrons with cal response to individual particles.• Pt spectrum of particles in jets varies slowly with Jet Pt.• EM calorimeter response is linear. The E/p for photons, π0, electrons in simulation are same as data within 1%. •The E/p of hadrons varies from ~0.55 to ~0.85.

Pythia PtMin=40

Jet HadPt

Particle P

Hadron Momentum (GeV)

Page 11: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Fraction of Jet Pt carried by particles with Pt<PtMax,

Track Pt Max (GeV)

<JetPt>=24 Gev 55 GeV

95 GeV 150 GeV

250 GeV 320 GeV

ΣPt (Pt<PtMax)/ΣPt(all)Jet 20,50,70, 100Tracks in cone R=0.7 around calorimeter jet axisNo tracking efficiency corrections

Pt Hadron-Jet

Fraction of Jet Pt carried by particles With Pt<400 MeV

A large fraction of jet energy is carried by low Pt particles

Fra

ctio

n of

Jet

Ene

rgy

Page 12: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Cal2Had Corrections

Generate DiJet PYTHIA Tune A events from PtMin=0,10,…, 600 GeV.

Reconstruct jets from HEPG stable particles using CDF clustering algorithms

Reconstruct jets from Calorimeter towers using same algorithm

Select Had-Jet, find the closet calorimeter jet within ηφ spece, require R<0.1.

Calculate the correction for CalorJet Pt to recover Hadron-Pt.

Cal-Pt distributions for fixed Had-Pt Jets

Z=CalPt/HadPt

8 GeV 20 GeV

50 GeV 200 GeVN

umbe

r of

Eve

nts

JetClu R=0.4

Page 13: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Systematic Uncertainty

E/p measurement/modeling P spectrum in jets (Tracking efficiency) Time dependence of data calibration (assume detector simulation is fixed

and agrees with data in limited time period.)

Page 14: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Calorimeter Calibration

Very good agreement up to 20 GeV.

For p> 20 relies on 1990 test beam. Test beam uncertainty ~2%. Not easy to check with data.It affects higher Pt jets.

Measurement limited to central 36% of the tower. Need to check φ/η crack simulation further.Sys Uncertainty 0 < p < 12 GeV 2%12 < p < 20 GeV 3% p> 20 GeV 4%

Track momentum (GeV)

EmHad

Em+Had

Em Objects agree to 1%

After background subtraction

E/p

Page 15: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Uncertainty from E/p calibration

Use GENP, stable particles within R=0.7 around hadron jet axis.

Calculate Cal Energy(E0) π0, photon, electron E/p=1.0 Charged particles using Soon’s E/p

curve on page 11. Change E/p ↑↓ and calculate E+/E- Plot E+/E0 and E-/E0Uncertainty0 < p < 12 GeV 2 %5 < p > 20 GeV 3 % p > 20 GeV 4 % (Run I-Run II understood to ~2%, 1990 Test Beam 2%, CHA response)Photon/π0/electrons <1%, ignore.Run I p<15 GeV 5% difference between in

situ calibration and test beam p>15 GeV 2% test beam calibration

Change in Jet Energy scale

PtMin 0,10,18,40,60,90,150,200,300,400,500

Page 16: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Fragmentation Uncertainty

If E/p was flat, uncertainty in Pt spectrum of particles in the jet will not lead to any uncertainty in energy scale.

Simon has compared P distribution in data and Pythia MC and show reasonable agreement except at low momentum particles for jets up to 100 GeV.

Need estimate the difference in Jet energy by keeping E/p fixed and changing P of particles.

HERWIG, wait and see. Photon Jet shows Pt spectrum is softer. How to assign systematic? Use HERWIG/Pythia difference as systematic?

Page 17: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Under-Reported Energy for Data and Pythia Tune A

Data Jet20,50,70,100Pythia Tune A 00,…500

Pythia “loss” 1% lower for Pt<250 GeV

Compare energy loss in calorimeter between in data and MC

Use tracks for MC and Data. Cone of R=0.7 around jet axis (should be small effect as E/p is

flat and large) Not sensitive to Pt<0.3 GeV

particles. (10% at 5 GeV, 1% Pt>50 GeV)

5% change in tracking efficiency 2% change effective E/p

Check HERWIG sample Need more checks but I think

uncertainty is ~1-2%

Loss (ΣPt- ΣPt●Response(p))/PtCalJet

Page 18: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Uncertainty of Out-of-Cone Corrections

W+Jet data HERWIG MC

Photon-Jet data

Pythia

HERWIG

Jet Isolation Cut , mainly gluons

Hard Δφ cuts/DiJet background not important/Quarks/gluons

0.5% uncertainty

Correcting back to parent partonCompare energy just outside the cone in data/Pythia/MC

New method: Compare fully corrected jet with photon

Page 19: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

DiJet Mass Resolution

Reorganization of jet corrections group, di-jet mass, zb\bar b groups

Build on Di-Jet mass group work (Stefano, Andrea from Rockefeller) Using tracking and ShowerMax/CPR information to correct the jet energies.

Tower Classification algorithm already in place. A simpler algorithm based only on calorimeter and tracks (H1) is also

working. Working on improvements by correcting towers before clustering (Andrea

Bocci wrote initial code). Further studies to evaluate and optimize the algorithms.

Page 20: Inclusive Jet Cross Section Jet Energy Corrections Anwar Ahmad Bhatti DOE Meeting December 2, 2004

Conclusions Not discussed in this talk, the CDF calorimeter is properly calibrated and stable after run

dependent calibrations. (Early in the run Plug calorimeter was found to be decaying very fast.)

We understand the origin of central calorimeter scale change from Run I to Run II. The energy scale uncertainty is substantially reduce from Gen 4 analysis.

New uncertainty is almost same as Run I, if not slightly better. Cal2Had Corrections

E/p measurement/modeling <2% (10 GeV) to 2.6% (600 GeV)

Fragmentation (need further studies/HERWIG) ~1-2% Relative and Multiple p\bar p subtraction uncertainty has also decreased. Out-of-Cone uncertainty 4.5—1%

Top mass papers soon. Corrections to finalized in next few weeks Need some more work for very high Pt Jets Refine/determine JetClu and MidPoint corrections (Kt clustering) Further investigate Pythia/Herwig fragmentation.

There is still room for improvements specially in simulation, relative corrections at low Pt, Out-of-cone corrections, and understanding HERWIG/PYTHIA differences.