1 / 32

Summary of Calorimeter PFA Status

Summary of Calorimeter PFA Status. Steve Magill Argonne National Laboratory. Motivation for Particle-Flow Jet Reconstruction Calorimeter Optimization for Particle-Flow Particle-Flow Algorithms Summary. Victoria LCWS 2004. Motivation for Particle-Flow Reconstruction.

ashby
Download Presentation

Summary of Calorimeter PFA Status

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Summary of Calorimeter PFA Status Steve Magill Argonne National Laboratory Motivation for Particle-Flow Jet Reconstruction Calorimeter Optimization for Particle-Flow Particle-Flow Algorithms Summary Victoria LCWS 2004

  2. Motivation for Particle-Flow Reconstruction Jets are important (necessary!) for precision measurements at the LC • Can explore EWSB thru the interactions : e+e- -> WW and e+e- -> ZZ • -> Requires Z,W ID • -> Can’t use (traditional) constrained fits W,Z Physics Requirement : separately id W, Z using dijet mass in hadronic decay mode (~70% BR)-> higher statistics physics analyses Detector Goal : measure jets with energy resolution -> /E ~ 30%/E 75%/M 30%/M

  3. A Particle-Flow Jet Detector • Jets : • charged pions, photons, KL0/n -> >95% of jet energy • other hadrons/leptons/neutrinos -> few % • Detector A – sensitive only to charged pions (+ other charged particles) • = Tracker • Detector B – sensitive only to photons • = ECAL-> try to force hadron showers into the HCAL • (longitudinal segmentation), fine granularity for separation of showers • Detector C – sensitive only to neutral hadrons • = HCAL-> longitudinal segmentation and fine granularity for separation of charged/neutral hadron showers

  4. Traditional Standards Hermeticity Uniformity Compensation Single Particle E measurement Outside “thin” magnet (~1 T) P-Flow Modification Hermeticity Optimize ECAL/HCAL separately Longitudinal Segmentation Particle shower reconstruction Inside “thick” coil (~4 T) Optimized for best single particle E resolution Optimized for best particle shower separation/reconstruction Particle-Flow Implications for Calorimetry 3-D shower reconstruction in ECAL/HCAL requires high degree of longitudinal segmentation and transverse granularity

  5. Common Particle-Flow Misconceptions • “Particle-flow algorithms can’t work for jets since particles overlap.” • -> almost always the result of 2-D thinking. Lots of calorimeter people are used to summing energy in towers (E-Flow) rather than reconstructing individual particle showers in 3-D (P-Flow). The starting point of a hadron shower in depth is very important to use in P-Flow. • 2) “Because of backgrounds (QCD), a jet resolution of 30%/E is physically unobtainable.” • -> comes from people used to hadron colliders where underlying event and FS gluon radiation dominates the jet energy resolution – clearly not the case at e+e- colliders. (see results)

  6. Results of Hadronization + Jet Algorithm Effects  Single Jet ep e+e-ttbar  e+e-ttbar  e+e-W+W- • Form of resolution fit : • /E ~ a/E  b/E • a ~ 0.1b For E > ~40 GeV, physics limits < P-Flow goal /E = 30%/E

  7. ECAL Requirements for Particle-Flow • Need a dense calorimeter with optimal separation between the starting depth of EM and Hadronic showers. If X0/I is small, then the longitudinal separation between starting points of EM and Hadronic showers is large • For electromagnetic showers in a dense calorimeter, the transverse • size is small • -> small rM (Moliere radius) • -> If the transverse segmentation is of size rM, get optimal transverse separation of electromagnetic clusters. • Some examples : • Material Z A X0/I • Fe 26 56 0.0133 • Cu 29 64 0.0106 • W 74 1840.0019 • Pb 82 2070.0029 • U 92 2380.0016

  8. ECAL Detector Designs CALICE Si/W, 5 mm -> 1 cm transverse size of cell U Oregon, SLAC

  9. Optimizing the HCAL for Particle-Flow • To optimize the HCAL for P-Flow requires : • full containment of hadronic showers. • good precision on energy measurement. • highly segmented in transverse and longitudinal directions in • order to separate in 3-D close-by clusters in jets. • Requires integrated approach which includes other detector sub-components in the design phase and incorporating P-Flow algorithm. • Assume a tracking system optimized for, e.g., di-lepton • measurements. • Assume an ECAL optimized for photon reconstruction. • Vary HCAL parameters, e.g., absorber material, thickness, • size of readout cells in both transverse and longitudinal • directions, to determine optimal performance in a P-Flow • Algorithm.

  10. Analog or Digital Hadron Calorimeter Readout? • Performance is function of energy and cell size • – as E  and cell size remains fixed, the density of tracks  • -> >1 crossing per cell • -> Cell size must be small • P-Flow : requires high transverse and longitudinal granularity • -> Cell size must be small • Digital calorimetry seems well suited • for P-Flow jet measurements

  11. Analog vs. Digital HCAL Linearity Response • Showers in a sampling calorimeters are characterized by their spatial development in terms of “track length” : • Track Length (T) = sum of tracks of all charged particles in a shower – Analog sampling calorimeters sum energy, Digital sampling calorimeters sum hits • T  E (particle energy) – what about spread in energy?

  12. Analog vs Digital Energy Resolution • GEANT 4 Simulation of SD Detector (5 GeV +) • -> sum of ECAL and HCAL analog signals - Analog • -> number of hits with 10 MeV threshold in HCAL - Digital Analog Digital Landau Tails + path length Gaussian /mean ~22% /mean ~19% E (GeV) Number of Hits

  13. SD Detector – a Particle-flow Detector for the LC • Tracking : • Multi-layer Si Vertex Detector • ~1 cm -> ~7 cm radius, 5 layers • Si-Strip Tracker • ~20 cm -> ~1.25 m radius, 5 layers • ECAL : • 30 layers, ~1.25 m -> ~1.40 m radius • W(0.25 cm)/Si(0.04 cm) • ~20 X0, 0.8 I • ~5 mm X 5 mm cells • HCAL : • 34 layers, ~1.45 m -> ~2.50 m radius • SS(2.0 cm)/Scin(1.0 cm) • ~40 X0, 4 I • ~1 cm X 1 cm cells • Solenoid Coil : • 5 Tesla, ~2.50 m -> ~3.30 m radius • Muon (Tail Catcher) : • ~3.40 m -> ~5.45 m

  14. Charged Particle Contribution at Z

  15. Particle-Flow Algorithm Approaches • Calorimeter Cluster-based Algorithms • -> start with calorimeter cell clustering ~ particle showers • Cluster ID by Neural Net • Many variables used to determine particle origin of cluster including tracking input • Weighted Calorimeter Clusters • Density or energy weights used to link calorimeter cells • Tracks matched to clusters – use track p • Sub-cluster ID • Separately cluster EM, mip, and hadronic parts of a particle shower • “perfect” compensation • “pixel” calorimeter • No tracking needed! • Track Extrapolation/Shower Association Algorithms • No calorimeter clustering needed – cell-by-cell association to extrapolated track with various algorithms • Leftover cells are photons (ECAL), neutral hadrons

  16. Track Extrapolation Particle-flow Algorithm ANL, SLAC 1st step - Track extrapolation thru Cal – substitute for Cal cells (mip + ECAL shower tube + HCAL tube; reconstruct linked mip segments + iterated in E/p hits in tubes) - analog or digital techniques in HCAL – Cal granularity/segmentation optimized for separation of charged/neutral clusters 2nd step - Photon finder - use analytic long./trans. energy profiles, ECAL shower max, etc. 3rd step - Jet Algorithm - tracks + photons + remaining Cal cells in jet cones defined by charged track jets (neutral hadron contribution) - Cal clustering not needed -> Digital HCAL?

  17. Shower reconstruction by track extrapolation ECAL HCAL • Mip reconstruction : • Extrapolate track through CAL layer-by-layer • Search for “Interaction Layer” • -> Clean region for photons (ECAL) • Shower reconstruction : • Define tubes for shower in ECAL, HCAL after IL • Optimize, iterating tubes in E,HCAL separately (E/p test) IL track shower

  18. CAL Interaction Layer G4v6.1 Single 5 GeV + All charged particles from e+e-  Z at 91 GeV

  19. Track-first P-flow Algorithm Results G4v6.1 • Average cone size as determined by the number of iterations : • ECAL – 0.12 rad • HCAL – 0.22 rad Start with green distribution -> end with yellow Lots of max iterations . . .

  20. Track Extrapolation/Shower Link Algorithm • 1. Pick up all seed cells close to extrapolated track • - Can tune for optimal seed cell definition • - For cone size < 0.1 (~6o), get 85% of energy • 2. Add cells in a cone around each seed cell • through n layers • 3. Linked seed cells in subsequent cones form • the reconstructed shower • 4. Discard all cells linked to the track Of course, neutrals are Non-Linked Cells

  21. Single 10 GeV Pion – event display comparison Blue – density = 1 Red – density = 2,3 Green – density > 3 Energy weight Density weight

  22. Track Substitution, Neutral Sum Results G4v6.1 • Jet cones – 0.5 • Neutral contribution to E sum ~3.7 GeV (most) • -> Goal is ~3 GeV (all) Includes mips + cell energies in conical tubes Further tuning of E/p parameter is still needed

  23. Total Energy Sum – Tracks + Photons + Neutrals G4v6.1 • 2 Gaussian fit : • mean = 83 GeV  = 7.8 GeV Poor hadron shower reconstruction with tubes – more sophisticated approach mean = 87 GeV • = 4.7 GeV Good hadron shower reconstruction

  24. Summary • Several complementary approaches to PFA development are proceeding worldwide. • -> ready for integration of realistic tracking detectors/pattern recognition algorithms. • -> opportunities for innovative ideas in calorimeter design and utilization. • From simulation studies, it appears that a Digital Hadron calorimeter is a particularly good choice for this component in a PFA approach : • -> good energy resolution and linearity. • -> small cell size optimized for shower reconstruction. • PFA algorithm development progressing towards goal of obtaining jet energy resolution of ~30%/E : • -> possible to obtain neutral contribution of ~3-4 GeV to Z mass in SD Detector. • -> proceed with better hadron shower reconstruction. • -> add photon finder, neutral cluster finder? • -> add real track reconstruction including forward tracking. • -> optimize detector using full PFA (SD -> LD, TESLA?). • -> understand relationship jet  BnRmaplq in PFA approach (B-field, Cal radius, Cal cell area, Cal longitudinal segmentation).

  25. CDF E-Flow ZEUS Jet E correction

  26. Optimization of Calorimeter Resolution • Hermeticity • Beampipe-to-beampipe coverage at colliders • Longitudinal segmentation • EMCAL/HCAL • Sandwich cals • Compensation • Hardware – precise absorber/detector thickness ratio • Software – finer segmentation – software separation of photon/hadron measurements • Energy-, Particle-flow approach

  27. Motivation for Track-First P-Flow Charged particles ~ 62% of jet energy -> Tracker /pT ~ 5 X 10-5 pT ~190 MeV to 100 GeV jet energy resolution Photons ~ 25% of jet energy -> ECAL /E ~ 15-20%/E ~900 MeV to energy resolution Neutral Hadrons ~ 13% of jet energy -> HCAL resolution not critical ~3 GeV to energy resolution • Also, since ECAL is dense, hadrons are optimally separated from photons (starting point of shower longitudinally) • -> 75% of hadrons shower after photon shower-max in ECAL

More Related