Track Trigger for Taus Michael We Michael W inberger inberger and Teruki and T Kamon In collaboration with V dim Vad Va i d m i m Khotilovich

Transcription

1 Track Trigger for Taus Michael Weinberger and Teruki Kamon In collaboration with Vadim Khotilovich and Alexei Safonov Many thanks to Laura Fields and Anders Ryd for help with their track stub code o o o o o o o OUTLINE Why taus? Phase-I Calorimeter Tau Trigger Phase-II Tracker Tau Trigger with Tracks Full Simulation Samples and Tools First Results Summary and Plan April 27,

2 Why Tau s? and Tau Trigger Physics with hadronically-decayingdeca ing taus in Higgs and SUSY Access for Z eventsents as the SM candle Calorimeter-based tau triggers at higher luminosity will be very difficult in an environment with huge QCD production cross sections with large amounts of overlapping pileup (PU) events. This requires that we will need to re-design and validate the trigger for both Phase I and II 2

3 Phase-I Calorimeter Tau Trigger Khotilovich, Mason, Safonov: Full calorimeter/muon simulation framework for simulating events in the high PU environment L1 Tau Tau(new) L1 Tau 40 GeV Tau L1 Tau(new) Trigger Rate for QCD events L1 Level 1 trigger 100 MHz Tau Tau (original config.) Tau(new) New tau trigger Tau L1 at PU50 Tau > L1 at PU100 Tau(new) degrading at PU100 We need further reduction of the trigger rate by requiring a track trigger primitive. 3

4 Phase-II Tracker Weinberger and Kamon - Full Simulation Package for Tracker (Long Barrel Geometry) - emphasis on improving tau triggering - in the SLHC Tracker Simulation Group 4

5 Stubs and Tracklets r-z view r-φφ view Stub Tracklet (vector!) 5

6 Current Long Barrel Geometry in 226 Phase 1 inner pixels In this study, the 4 th inner pixel is not included. 6

7 Tau Trigger with Tracks Assuming the matching of the seed track and a cluster can be done like in electron trigger, our focus is to test several isolation (ISO) quantities: 1) N sub =[# of stubs in ISO cone for layer k] 2) N = [# of tracklets in ISO cone] 3) I = [Σp T for tracklets in ISO cone] 4) I θ = Σ ( p T θ ) / Σp T 5)... where ISO cone = between R S and R I R S = Singnal cone radius R I = Isolation cone radius I 7

8 Full Simulation Samples and Tools (1) Full simulation package for Long Barrel Tracker (M. Weinberger) + Calorimeter (V. Khotilovich) in 226 ½ detector (to offset memory problems) (2) Test samples Ditau (p T = 10~200 GVη GeV, η = 0~2.5) + PU0 Ditau (p T = 10~200 GeV, η = 0~2.5) + PU200 [2900 events] (3) Analysis codes Laura Fields s track stub finder in 226; Clustering of stubs to remove mutiple counting not ready. Michael Mason s Phase-1 tau calo-cluster reconstruction in under migration to 226 (not fully tested yet) Today, we focus on N sub. 8

9 Hits First Look: Hits and Stubs Hit Position Stub Position y (cm m) r (cm m) η = 0 Stubs Layer Number x (cm) Stub Position z (cm) Stub Position y (cm m) Layer Number x (cm) z (cm) 9

10 First Look: Stub Rates (PU200) MHz z/cm 2 However these rates count every pixel for use in stub creation. So many stubs will be created from a single track Need clustering to remove the extra stubs Layer Number 10

11 Crude Calculation: N stub (1) Find the average rate of subs per area by layer (2) Find area of ISO cone ( θ =10 o ~ 30 o ) for a track at 45 o. (3) Combine to find the average number of stubs in the ISO cone. N stub =Area*2*Rate/40MHz 11

12 N stub in 10 o -30 o Cone one o -30 o C b in 10 o N stub Layer Number 12

13 N stub in 10 o Cone Can not compare stubs to simstubs as Fullsim does not have PU in sim path Fastsim does, so look at 50pu events and see how many stubs versus the simstubs to get estimate of overcounting Factor F t of ~10 13

14 Summary and Plan (1) We created a full simulation package for Long Barrel Tracker + Calorimeter in 226 and tested a fully simulated edsample of Ditau +PU200 events. (2) We plan to have : A full production sample of QCD events A complete analysis package Laura Fields s track stub finder in 226 Clustering to be prepared. Michael Mason s tau cluster reconstruction in migration to 226 A complete performance evaluation Isolation Track-clustermatching cuse c 14

15 Backup 15

16 Calculation: Nstub in 10o-30o Cone 16