LCD Simulation News Norman Graf ECFA Workshop ECFA Workshop Montpellier, Nov. 14, 2003

LCD Simulation Overview Providing canonical data samples for LC community. Providing canonical data samples for LC community. Aim for more realistic simulation of detector response. Aim for more realistic simulation of detector response. Better geometry descriptions in GEANT4. Better geometry descriptions in GEANT4. Digitization of resulting hits in detectors. Digitization of resulting hits in detectors. Include all backgrounds (machine & physics) Include all backgrounds (machine & physics) Developing more functional fast MC for rapid detector prototyping. Developing more functional fast MC for rapid detector prototyping. Modularizing and improving reconstruction. Modularizing and improving reconstruction. Encourage collaboration! Encourage collaboration!

Simulation SLAC 4 day workshop devoted to simulation tools. 4 day workshop devoted to simulation tools. Two days JAS/AIDA/Wired/Geant4 Two days JAS/AIDA/Wired/Geant4 Two days lcd-specific software Two days lcd-specific software AM presentations / PM tutorials AM presentations / PM tutorials Two dozen participants Two dozen participants www-conf.slac.stanford.edu/lcsimworkshop/ www-conf.slac.stanford.edu/lcsimworkshop/ www-conf.slac.stanford.edu/lcsimworkshop/ Would like to schedule another such workshop if interest exists. Would like to schedule another such workshop if interest exists. Who? Where? When? Who? Where? When?

Data Samples - Backgrounds Large sample of beam backgrounds available. Large sample of beam backgrounds available. e + e - pairs, photons, muons. e + e - pairs, photons, muons. Full  hadrons sample being generated Full  hadrons sample being generated beamstrahlung-beamstrahlung - done beamstrahlung-beamstrahlung - done beam+bremsstrahlung & brem-brem – ongoing beam+bremsstrahlung & brem-brem – ongoing Run through full detector simulations. Run through full detector simulations. All samples can be overlaid onto signal events at the detector hits level. All samples can be overlaid onto signal events at the detector hits level. Can adjust level + mix of backgrounds Can adjust level + mix of backgrounds

Data Samples - Signal 500GeV 2ab -1 SM sample being generated 500GeV 2ab -1 SM sample being generated will complement the existing sample at 1TeV. will complement the existing sample at 1TeV. aimed at fast MC analyses. aimed at fast MC analyses. only subsamples will be fully simulated. only subsamples will be fully simulated. Signal samples: Signal samples: WW, ZZ, Zh, Zhh, Z , tt, qq, , , SUSY(SPS1)… WW, ZZ, Zh, Zhh, Z , tt, qq, , , SUSY(SPS1)… Full detector simulations. Full detector simulations. JAS server at lcddata01.slac.stanford.edu JAS server at lcddata01.slac.stanford.edu ftp://ftp-lcd.slac.stanford.edu/lcd/NewData/ “Single particle” diagnostic samples.

Intermediate Fast Simulations Fast Monte Carlo systems which generate hits in trackers and calorimeters. Fast Monte Carlo systems which generate hits in trackers and calorimeters. Used for rapid prototyping of detector configurations. Used for rapid prototyping of detector configurations. Allow pattern-recognition questions to be answered within a fast MC framework. Allow pattern-recognition questions to be answered within a fast MC framework. Lelaps (W. Langeveld) Lelaps (W. Langeveld) Java shower parameterizations (S. Pathak) Java shower parameterizations (S. Pathak)

Lelaps Standalone C++ library & program. Standalone C++ library & program. Internal diagnostic event generator or stdhep input. Internal diagnostic event generator or stdhep input. Multiple scattering and dE/dx for tracks. Multiple scattering and dE/dx for tracks. Photon conversions and V decays. Photon conversions and V decays. Parameterizations for calorimeter showers. Parameterizations for calorimeter showers. EM showers for e, . EM showers for e, . MIP in ECAL + had. showers for charged hadrons. MIP in ECAL + had. showers for charged hadrons. MIP traces in all calorimeters for muons. MIP traces in all calorimeters for muons. SDJan03 and LDMar01 detectors implemented. SDJan03 and LDMar01 detectors implemented. Other configurations can easily be added. Other configurations can easily be added. SIO output, LCIO in beta. SIO output, LCIO in beta.

Lelaps event e + e -  Zh  bb

Full Simulation LCDG4 producing sio output, lcio soon. LCDG4 producing sio output, lcio soon. sdjan03, ldmar01 sdjan03, ldmar01 simplified TESLA detector for comparisons. simplified TESLA detector for comparisons. xml detector descriptions, easily modified. xml detector descriptions, easily modified. American xml detector geometry description also available within Mokka framework. American xml detector geometry description also available within Mokka framework. Complex beamline geometries added Complex beamline geometries added LCIO output available LCIO output available QA ongoing QA ongoing Merger of two desirable! Merger of two desirable!

LCIO Simple, extensible data model. Simple, extensible data model. European & American standard for IO. European & American standard for IO. Persistence uses, but not tied to, SIO. Persistence uses, but not tied to, SIO. Java and C++ interface (Fortran also). Java and C++ interface (Fortran also). v1 released! v1 released! Implemented in Mokka & lelaps. LCG4 soon. Implemented in Mokka & lelaps. LCG4 soon. See Frank Gaede’s talk for details. See Frank Gaede’s talk for details.

Persistent Data: MC Event definition and framework. Event definition and framework. MC output formats defined for “generic” tracker and calorimeter hits. MC output formats defined for “generic” tracker and calorimeter hits. SimTrackerHits for position-sensitive detectors SimTrackerHits for position-sensitive detectors SimCalorimeterHits for energy-sensitive detectors. SimCalorimeterHits for energy-sensitive detectors. Monte Carlo Particle heirarchy classes. Monte Carlo Particle heirarchy classes. Formalism exists for handling secondaries produced in full simulations. Formalism exists for handling secondaries produced in full simulations. Have both generator and simulator status words. Have both generator and simulator status words.

Persistent Data: Reconstruction Tracks & Calorimeter Clusters Tracks & Calorimeter Clusters Requirements well understood Requirements well understood First implementation ~written First implementation ~written Use cases being assembled Use cases being assembled ReconstructedParticle ReconstructedParticle Requirements still need to be fleshed out. Requirements still need to be fleshed out. Hope to initiate discussion of requirements. Hope to initiate discussion of requirements. Persistent implementation not yet available. Persistent implementation not yet available.

Detector Response I Emphasis being placed on more realistic detector response. Emphasis being placed on more realistic detector response. Replace MC hit smearing with full chain, e.g. Replace MC hit smearing with full chain, e.g. MC “hits” in VertexDetector  CCD Pixels MC “hits” in VertexDetector  CCD Pixels Associate CCD Pixels into clusters Associate CCD Pixels into clusters Extract hit positions & uncertainties from clusters. Extract hit positions & uncertainties from clusters. VXD Hits  Pixels based on SLD data (N. Sinev) VXD Hits  Pixels based on SLD data (N. Sinev) Si Hits  Strips: MC (C. Flacco, B. Schumm) Si Hits  Strips: MC (C. Flacco, B. Schumm) Calorimeter MC Hits digitized  ADC hits. Calorimeter MC Hits digitized  ADC hits.

Detector Response II Currently noise and inefficiencies added at the “hit level”, replace with more realistic readout channel noise and inefficiencies. Currently noise and inefficiencies added at the “hit level”, replace with more realistic readout channel noise and inefficiencies. Digitization will allow effects of hit-merging & ghosting to be systematically studied. Digitization will allow effects of hit-merging & ghosting to be systematically studied. Could also study effects of readout thresholds, dead channels, crosstalk, etc. Could also study effects of readout thresholds, dead channels, crosstalk, etc. More immediately useful for simulation of testbeam setups and readout. More immediately useful for simulation of testbeam setups and readout.

Event Reconstruction Framework LCDEvent model being refactored to improve design and modularize components. LCDEvent model being refactored to improve design and modularize components. Analysis examples being modified to access LCIO data instead of SIO. Analysis examples being modified to access LCIO data instead of SIO. Aiming for analysis-framework-independent reconstruction capability. Aiming for analysis-framework-independent reconstruction capability. Standalone reconstruction program produces lcio output and/or histograms/ntuples. Standalone reconstruction program produces lcio output and/or histograms/ntuples. Analysis proceeds in framework of choice. Analysis proceeds in framework of choice. Can also do everything within Integrated Analysis Environment. Can also do everything within Integrated Analysis Environment.

Reconstruction Track finding/fitting being implemented for axial-only SiD tracker and forward tracker configurations. Track finding/fitting being implemented for axial-only SiD tracker and forward tracker configurations. Calorimeter cluster analysis package available Calorimeter cluster analysis package available Several analyses targeted at a full “Energy Flow” reconstruction underway. Several analyses targeted at a full “Energy Flow” reconstruction underway. Aim to characterize detector performance in terms of resolutions, efficiencies & fake rates based on full ab initio reconstruction of digitized (“raw”) hits in presence of noise. Aim to characterize detector performance in terms of resolutions, efficiencies & fake rates based on full ab initio reconstruction of digitized (“raw”) hits in presence of noise.

Documentation & Organization Tutorials and other documentation being updated as code base changes. Tutorials and other documentation being updated as code base changes. Attempting to remove/improve outdated or incorrect webpages. Attempting to remove/improve outdated or incorrect webpages. Need feedback from users! Need feedback from users! What works What works What doesn’t What doesn’t Encouraging use of cvs for archiving and storing software/documentation. Encouraging use of cvs for archiving and storing software/documentation. Instituting formal testing + release schedule. Instituting formal testing + release schedule.

Status Simulation & Reconstruction efforts are currently in a state of flux: Simulation & Reconstruction efforts are currently in a state of flux: GISMO/SIO transitioning to Geant4/LCIO. GISMO/SIO transitioning to Geant4/LCIO. JAS2  JAS3 JAS2  JAS3 Smeared MC Hits  Digitized readout Smeared MC Hits  Digitized readout Using this occasion to refactor many aspects of the reconstruction packages. Using this occasion to refactor many aspects of the reconstruction packages. Aim for seamless transition. Aim for seamless transition. International collaboration welcomed & encouraged. International collaboration welcomed & encouraged.

Continuing Communication Mailing lists: lcd-sim, lcd-dev Mailing lists: lcd-sim, lcd-dev SIM/subscribe.html SIM/subscribe.html SIM/subscribe.html SIM/subscribe.html DEV/subscribe.html DEV/subscribe.html DEV/subscribe.html DEV/subscribe.html lcnews HyperNews forum lcnews HyperNews forumHyperNews forumHyperNews forum Regular phone meetings (esnet?, VRVS?) Regular phone meetings (esnet?, VRVS?) New website being developed New website being developed Plan to start with fresh portal; only current information Plan to start with fresh portal; only current information