1P. LochP. Loch

U of ArizonaU of Arizona

October 13, 2009October 13, 2009Radial Smearing

Simple radial energy distribution in tower Simple radial energy distribution in tower gridsgrids Ignore longitudinal developmentIgnore longitudinal development

Particle energy distributed transverse to direction of flight of Particle energy distributed transverse to direction of flight of particle in a plane through the particle impact point into the particle in a plane through the particle impact point into the calorimetercalorimeter

Shape of distribution from experiment/full simulationShape of distribution from experiment/full simulation

Integrated energy in profile is the same as particle energyIntegrated energy in profile is the same as particle energy No calibration/acceptance/smearing of energyNo calibration/acceptance/smearing of energy

Distributed energies projected into regular eta/phi grid Distributed energies projected into regular eta/phi grid within modeled detector acceptancewithin modeled detector acceptance Fakes calorimeter tower signal definition, including high eta Fakes calorimeter tower signal definition, including high eta

losseslosses

Different grid and eta acceptance for EM particlesDifferent grid and eta acceptance for EM particles

2P. LochP. Loch

U of ArizonaU of Arizona

October 13, 2009October 13, 2009Modeling

Detector defaultsDetector defaults EM acceptance -2.5 < eta < 2.5EM acceptance -2.5 < eta < 2.5

Photons/electrons outside are mapped onto HAD towersPhotons/electrons outside are mapped onto HAD towers HAD acceptance -5.0 < eta < 5.0HAD acceptance -5.0 < eta < 5.0

Particles outside are ignored completelyParticles outside are ignored completely Cylindrical calorimeterCylindrical calorimeter

R = 1200 mm, -2500 mm < z < 2500 mmR = 1200 mm, -2500 mm < z < 2500 mm High granularityHigh granularity

0.025 x 0.025 (EM)0.025 x 0.025 (EM) 0.1 x 0.1 (HAD)0.1 x 0.1 (HAD)

Shower shapesShower shapes Presently Gaussian within cylinderPresently Gaussian within cylinder

Lateral extend /cylinder radius 80 mm for EM particles (also in Lateral extend /cylinder radius 80 mm for EM particles (also in HAD grid!)HAD grid!)

160 mm for all others160 mm for all others Gaussian showers are too wide, this is just a simplification!Gaussian showers are too wide, this is just a simplification!

Energy distributed in small “spots”Energy distributed in small “spots”

3P. LochP. Loch

U of ArizonaU of Arizona

October 13, 2009October 13, 2009

Two files needed:Two files needed: DetectorModel.hDetectorModel.h and and DetectorModel.cxxDetectorModel.cxx

Uses Uses fastjet/PseudoJet.hfastjet/PseudoJet.h and and STLSTL – nothing else! – nothing else! Compile Compile DetectorModel.cxxDetectorModel.cxx

Usage (default):Usage (default): #include “DetectorModel.h”#include “DetectorModel.h” …… // default smearing module// default smearing module DetectorModel::RadialSmearing smear;DetectorModel::RadialSmearing smear; … … fill fill std::vector<PseudoJet> input std::vector<PseudoJet> input with final state particles and with final state particles and set set

thethe user_indexuser_index of each of those to the particle pdg code!!of each of those to the particle pdg code!! // retrieve pseudo-particles after smearing// retrieve pseudo-particles after smearing const std::vector<PseudoJet>& smeared = smear.smear(input);const std::vector<PseudoJet>& smeared = smear.smear(input); // do your thing…// do your thing…

Observations:Observations: Expect about 20 x more particles than in the generator final stateExpect about 20 x more particles than in the generator final state

Topology dependent, here Pythia QCD di-jetsTopology dependent, here Pythia QCD di-jets Slow executionSlow execution

Needs much more optimizationNeeds much more optimization

Software

4P. LochP. Loch

U of ArizonaU of Arizona

October 13, 2009October 13, 2009Sanity Checks

5P. LochP. Loch

U of ArizonaU of Arizona

October 13, 2009October 13, 2009Sanity Checks

6P. LochP. Loch

U of ArizonaU of Arizona

October 13, 2009October 13, 2009Sanity Checks

7P. LochP. Loch

U of ArizonaU of Arizona

October 13, 2009October 13, 2009Electron Lateral Profile (H1)

Points/open histogram: data + simulated signal with noise cuts

Shaded histogram: simulated shower development, no noise or cuts

PL Thesis, 1992

8P. LochP. Loch

U of ArizonaU of Arizona

October 13, 2009October 13, 2009Hadronic Lateral Profile (H1)

Points/solid line histogram: data + simulated signal with noise cuts

Dashed line histogram: simulated shower development, no noise or cuts

PL Thesis, 1992

9P. LochP. Loch

U of ArizonaU of Arizona

October 13, 2009October 13, 2009Improvements

Short termShort term Better radial profilesBetter radial profiles More segmentation regions More segmentation regions Better treatment of transitionsBetter treatment of transitions

Long termLong term Energy resolution smearingEnergy resolution smearing Hadron EM/HAD energy sharing (?Hadron EM/HAD energy sharing (? Magnetic fieldMagnetic field

Non-pointing (in azimuth) particle trajectory at impact point!Non-pointing (in azimuth) particle trajectory at impact point! Requires different treatment of phi in cylinder/cone Requires different treatment of phi in cylinder/cone

intersection intersection