1. TeV Muon Reconstruction Vladimir Palichik JINR, Dubna NEC’2007 Varna, September 10-17, 2007

2. O u t l i n e Motivation Muon reconstruction in CMS: local, regional and global Problems with high Pt muon reconstruction Current status of TeV muon reconstruction

3. Motivation CMS physical program in particular includes the searching of new TeV-mass resonances in dimuon channel and the testing of Standard Model prediction for the Drell-Yan continuum up to TeV energy scale studies with hard *) muons at CMS can lead to new physical phenomena discoveries and also be a tool for testing of reconstruction algorithms on their efficiency and a precision of measurements precision of measurements in the muon system plays a leading role in achieving of the required resolution on transverse momentum and mass reconstruction for the processes with hard muons Dubna group is responsible for ME1/1 production ME1/1 is the most precise station in the muon endcap system: (ME1/1) : (ME2  4) ~ 1 : 2 a special interest to high-p T muons (~ 1 TeV), where a role of the muon system increases *) hard muons – the muons with high transverse momenta from several hundred GeV up to TeV region

4. Bremsstrahlung and e-m showering closed to hard muons lead to the problems with reconstruction in the muon system

5. Muon Reconstruction Muon Reconstruction Local pattern recognition: Local Regional Global (individ.detectors) (Muon system) (Muon & Tracker system) 1.DT: time-space conversion of Digis *) into RecHits; CSC: clusters of Digis fitted into RecHits, i.e. position of hits in the detector layers. 2.Track-segment building (track following & track road methods) with the compatible RecHits in muon chambers. The track-segments are selected by a goodness of fit criteria (  2 ).  ~ 1-2% of strip width, i.e. 150-300  *) “Digis” – digitized electronic signals CSC

6. 1.CSC clusters of Digis fitted into RecHits 2.CSC track-segment building (track road method) with RecHits array in muon chamber: -start with the first and the last RecHits and build a straight line from them; -collect hits in a narrow road around the line (usually at least 4 hits ); -the track-segments are selected by a goodness of fit criteria (  2 ); -“used” hits are excluded for the next iteration; -if number of track-segments < 2 in a narrow road, try to build them in a wide road. CSC segment building Narro w road Wide road SK Algorithm of Local pattern recognition in CSC: Specifics for TeV muons: - high multiplicity of RecHits; - distortion of muon signals by  rays 5 algorithms for segment building in the CSCs in the current release of CMSSW – due to MTCC data handling

8.  = 118  m ME1/1 (top) resolution with MTCC2 data (results by CMSSW_1_2_2) 1)CSCRecHitB package with Calibration DB is used to find the RecHits; 2) make Gatti distribution more broader; 3)SK algorithm for CSC track-segments building; 4) requires 5,6 hits on a track segment; 5) cut on large angles of muons: dx/dz < 0.25 dy/dz <1.5 6)   < 6 cut for track-segments MTCC – Magnetic Test & Cosmic Challenge held in the 2006 year at CERN Resolution of the ME1/1 upper part is 118 microns

9. Regional Muon Reconstruction Standalone Muon system (DT,CSC and RPC): I Y Station 1 Station 3 Station 2 Station 4 e.m.showers& punchthrough Multiple scattering, energy losses, bremsstrahlung, e + e - pair production etc. in Iron Yokes (I Y) Recursive track fitting (from hit to hit in endcap; with use segments in barrel) by the least squares method ( Kalman filter propagation & parameter estimation) HCALHCAL from Hadron Calorimeter (HCAL) Specifics for TeV muons: - distorted RecHits - concurrent track-segments in a closed region

11. Global Muon Reconstruction inclusion of tracker hits starts from standalone reconstructed muons, propagated through calorimeters to outer tracker surface creating seed(s) for each propagated muon in (  )-window regional track reconstruction is performed using Tracker hits within this (  )- region; to resolve the ambiguities, all the reconstructed tracks are then refitted including the tracks in the muon chambers significantly improves Pt resolution – by factor of 10 for low Pt muons (comparing with standalone muon reconstruction) Specifics for TeV muons: hard muons look like straight lines in the Tracker - -> muon momentum can be reconstructed correctly only in the total tracking (Muon+Tracker) system

13. 1 TeV Muon Reconstruction in current CMSSW release Transverse momentum resolution = 9.6% Efficiency of reconstruction = 97.6% Endcap region 1.2 <  < 2.1

14. Progress with Single Pt = 1 TeV Muon Reconstruction in endcap (1.2 <  < 2.1) region 100 % 95 % 90 % 85 % efficiency ORCA releases 6.3.07.3.07.6.18.0.18.2.08.4.0 200220032004 dPt/Pt 15 % 10 % release - HLT - GMR 80 % 2005 8.7.38.13.2 release 2007 1.6.0 CMSSW release

15. Drell-Yan di-muon Mass Reconstruction in the previous CMS software (ORCA) (off-line simulation)  =94 %  ( dM/M ) = 4.0 % sample with di-muon invariant mass cut-off 1 TeV

16. Summary A significant progress in efficiency and transverse momentum reconstruction for single hard muons in the CMSSW: for Pt=1TeV the efficiency of reconstruction in the Endcap region is 97.6% & Pt-resolution stays less than 10% A significant progress in efficiency and transverse momentum reconstruction for single hard muons in the CMSSW: for Pt=1TeV the efficiency of reconstruction in the Endcap region is 97.6% & Pt-resolution stays less than 10% Consequently it is expected to obtain in this CMSSW version (1.6.0) mass resolution for di- muon processes in the TeV mass region Consequently it is expected to obtain in this CMSSW version (1.6.0) mass resolution for di- muon processes in the TeV mass region First MTCC data with cosmic muons gave a possibility to verify and adapt CMSSW to real data handling (calibrations, algorithms etc.) First MTCC data with cosmic muons gave a possibility to verify and adapt CMSSW to real data handling (calibrations, algorithms etc.)

17. Thank you for your attention

18. Single hard muon Reconstruction Single hard muon Reconstruction in 4 eta-regions (off-line)  =97 %  =94.5 %  =94 %  =96.5 %  ( dPtinv/Ptinv ) = 6.2 %  ( dPtinv/Ptinv ) = 5.2 %  ( dPtinv/Ptinv ) = 10.8 %  ( dPtinv/Ptinv ) = 32 % |  | < 0.8 barreloverlap endcap 1.2<|  | < 2.1 0.8 <|  | < 1.2 endcap |  | >2.1