Lecce Trigger in ATLAS Andrea Ventura INFN Lecce

Lecce Trigger in ATLAS Andrea Ventura INFN Lecce & Università del Salento V ATLAS-Italia Physics Workshop Napoli, 18 -19 May 2011

Outline • ATLAS Trigger strategy and challenges • Physics motivations and proposals for menus – Standard Model – Higgs, SUSY and Exotics • Trigger menus – Baseline Trigger and evolution toward higher luminosities • Trigger status, performance, rates and issues – Examples: Muons, Egamma • Involvement of Italian community in ATLAS Trigger – Muon L 1: Barrel performance and RPC studies – Muon HLT: algorithms, performance, validation – b-jet trigger • Conclusions 18/05/2011 A. Ventura - Trigger in ATLAS 2

ATLAS Trigger design LHC: 40 MHz • L 1 – hardware based – Max latency: 2. 5 s – Fast custom electronics – Define Regions of Interest (Ro. I) L 1: 75 k. Hz • L 2 – PC farm ( 500 nodes) – Mean processing time: 40 ms – Fast reconstruction within a Ro. I L 2: 3 k. Hz • EF – PC farm ( 1800 nodes) EF: 200 Hz Tier-0: 300 MB/s 18/05/2011 – – Mean processing time: 4 s Offline-like algorithms Ro. I-based or Full-Scan strategy Full event access, complete calibration and alignment data A. Ventura - Trigger in ATLAS Ro. I mechanism 3

Trigger: definition and strategy • Trigger is configured via a trigger menu, menu defined in chains, chains each one starting from L 1 with a sequence of FEX/ FEX HYPO for reconstruction/selection steps. • Caching allows to re-use features from a chain to another. • Menus are designed to maximize the physics potential – There are 256 L 1 items and up to 500 HLT chains – Triggers can be: single, di/multi-object, combined, topological – Categories: primary (unprescaled), support, calibration, monitoring • Constraints: 3 k. Hz @EF input, 300 MB/s @Tier-0, 200 Hz on tape • Higher luminosities imply: – Tighter object definitions (and isolation) against higher pile-up – Higher threshold for primary triggers – Higher prescales at lower p. T threshold for other triggers ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 4

Trigger for Standard Model • Minimum bias, underlying event, Soft QCD : minbias trigger – Low luminosity: detector studies unbiased by trigger, MC tuning, … • Jet physics: single/multi- jet triggers – Differential jet cross-sections, multiplicity, JES & resolution, calibration, … • Photon physics: photon triggers (only loose planned, no tight) – Differential photon & di-photon cross section, background studies, … • Electro. Weak physics, Drell-Yan: electron and muon (+ ETmiss) triggers – W/Z, mass, cross-sections, F/B asymmetries, di-bosons, constrain PDFs at low masses, … – Tag&Probe Z ll measurements, (di-)lepton threshold increasing with luminosity • ttbar analyses: lepton / jets / tau / b-jets triggers – Top mass, cross-section, properties; channels: ttbar l. X , ttbar h. X, ttbar X h. X – Examples: e 20_medium 1, mu 20, 2 e 12, 2 mu 10, b 10_tight_4 L 1 J 10, 4 j 55 • B physics: J/psi , Upsilon B triggers (topological, Ro. I-based, Full. Scan) – J/ , : Resonances, cross-section and polarization studies; low p. T triggers with invariant mass required in [2. 5, 4. 3] Ge. V and in [8, 12] Ge. V (e. g. 2 mu 4_Di. Mu) – B + -(X): mixing angles, CP violation, rare decays, … (e. g. 2 mu 4_Bmumu. X) ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 5

Trigger for Higgs, SUSY, Exotics • Higgs boson searches: photon / (di-)lepton / tau / b-jets triggers – Possible channels: , ZZ 4 e, 2 e 2 , 4 , WW, ; also Higgs in MSSM and NMSSM • Diphoton channel with 2 g 20_loose, or g 30_g 20_loose if needed • Single lepton plateaux expected at 22 -23 Ge. V; dileptons: 2 e 12, 2 mu 10, or also mu 15_mu 10_EFFS, e 10_medium_mu 6. • SUSY searches: – 0 leptons: multi-jets / jets+ETmiss / jet sum / b-jet triggers (j 75_xe 45, topological, . . ) – 1 or more leptons: electrons and muons (+ ETmiss) triggers (lowest unprescaled) – Pointing/non-pointing photons: photon / HV triggers (e. g. : g 40_loose_EFxe 40) – (Very) long lived massive particles: ETmiss / mu 40_MSonly* / mu_*slow / J*EMPTY • Exotics analyses: – Large variety of triggers (and for corresponding control samples too) • • • Graviton to two photons: 2 g 20_loose (di-)photon Mono-jets: jets, jets j 250 is lowest unprescaled; also ETmiss required Z’ ll, (di-)lepton / (di-)tau at high momenta; also from “boosted tops” Prompt lepto-jets: high multiplicity and small angles required (lepton+)jets Leptons from long lived particles (displaced vertex, or Hidden Valley), trileptons, 4 th generation production, Technicolor, and many others… ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 6

Present luminosity and online rates • Presently (May 2011), luminosity has reached 8. 8· 1032 cm-2 s-1, 700 bunches • LHC aims at 2· 1033 in short time, beams with up to 1400 bunches • Physics EF rates near 300 Hz: main streams include: Egamma (100 Hz), Muons (80 Hz), Jet. Tau. Etmiss (120 Hz) + Min. Bias, Background, Express… • Trigger rate reduction planned via various strategies (see in the following) ATLAS recorded 0. 27 fb-1 ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 7

Baseline trigger menu @1033 • Based on 2010 data, rates have been extrapolated to higher luminosities • Besides primary, primary TDAQ constraints have to be respected @1033 and beyond including support/ prescale, prescale priority 1 and May 3 2 additions + items Apr 28 with high L 1 rate, rate all Apr 24 scaling with lumi. Apr 19 • Also priority 1 triggers with fixed prescale 2010 • Presently discussing on possible increase of Predictions DAQ resources to get using 2010 to 300 Hz (with priority enhanced bias. 1) or 400 Hz (with also Pile-up effects not considered priority 2) Proposal & motivations: ATL-COM-DAQ-2011 -007 ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 8

Partition of allocated bandwidth • Each trigger group is given a fraction of the total EF output rate ( 200 Hz) according to the physics channels of interest • Increasing focus is on (multi) lepton triggers to improve possibility of discovery • Less room for first physics program or calibration • 1033 rates here are guestimates (depend on total EF output rate) ATLAS Italia, Napoli 18/05/2011 Trigger signature 1030 cm-2 s-1 1031 cm-2 s-1 1032 cm-2 s-1 1033 cm-2 s-1 Rate (Hz) Minimum bias 20 10 10 10 Electron/Photon 30 45 50 100 -150 Muon 30 30 50 50 -100 Tau 20 20 15 15 -30 Jet & forward jet 25 25 20 20 -30 b-jet 10 15 10 10 -20 B-physics 15 15 10 10 -20 ETmiss and ET 15 15 10 10 -20 Calibration triggers 30 13 13 13 A. Ventura - Trigger in ATLAS 9

Primary triggers at 1033 • List of main unprescaled triggers for physics (estimates are continuously updated!) • EF rates are peak rates (i. e. at start of fill; LHC refill when L=0. 5*Lpeak) • Additional 30 Hz include more specialized or “exotic” triggers. • Another 35 Hz is allocated for support/monitoring • This menu should survive until 1. 5· 1033, and baseline up to 2· 1033 ATLAS Italia, Napoli 18/05/2011 Evolution of straw-man menu A. Ventura - Trigger in ATLAS 10

What beyond 1033? • Trigger rate issues as the luminosity will increase: Even tighter cuts (e. g. 25 Ge. V on leptons) Removal of analysis-specific triggers Reduction of safety margin for EF output rate Algorithms to be tested against pileup (just guesses so far) • Priority 1 and Priority 2 lists have been compiled (see next) – In case of 400 Hz limit: • At 1. 5· 1033 drop Priority 2 400 Hz limit • At 2· 1033 drop Priority 1+2 – In case of 200 Hz limit: • Drop Priority 1+2 already 200 Hz limit – – after 1033 and tighten cut of the baseline menu — Constraints also from L 1/L 2 rate and deadtime • L 1 deadtime <1% at 75 k. Hz ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 11

PRIORITY Priority 1 triggers Proposal & motivations: ATL-COM-DAQ-2011 -007 ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 12

PRIORITY Priority 2 triggers Proposal & motivations: ATL-COM-DAQ-2011 -007 • In lower priority there also triggers with high L 1 rate – e. g. : e 15_tight, 2 e 10_medium. Tests needed, case by case. • At End-of-Fill some “low-luminosity” triggers can be used – For instance those which require low pile-up environment, high track multiplicity, large L 1 or L 2 bandwidth or latency ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 13

EF rate evolution vs. p. T thresholds • EF rates estimated at 1033 versus the tightness of trigger cuts on p. T thresholds • Most of rate should be due to high p. T leptons satisfying important final states with W/Z (SM, Higgs, ttbar, di -bosons, etc. ) • Dilepton rates are expected to be under control ATLAS Italia, Napoli 18/05/2011 electrons muons jets ETmiss A. Ventura - Trigger in ATLAS 14

Evaluating performance • Mostly data-driven methods to compute efficiencies, resolution, fake rates. – – Use of Tag & Probe for lepton studies (from J/ , , Z and from W for electrons) Use of orthogonal triggers (other streams with unbiased selection) Use of bootstrap method (study of lower threshold selections) Comparison data/MC • Estimates of efficiency scale factors / variable smearing / offsets and biases • Improvement of simulations (including more realistic geometry, inefficiencies, etc. ) • Offline objects defined in Combined Performance Groups as common reference for trigger studies – Need of updates as data are collected, and uniformity among Physics Groups • Performance of the ATLAS Trigger in 2010 running at the LHC • http: //cdsweb. cern. ch/record/1325333 , CDS internal note • Performance of the ATLAS Trigger System in 2010 – http: //cdsweb. cern. ch/record/1351486 , to be submitted to Eur. Phys. J. C ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 15

Muon Trigger status • Primary single muon trigger: mu 20 (mu 18 still unprescaled until 2· 1033) – mu 15 i isolation to allow lower thresholds (e. g. h/Z ) • Primary for di-muons: 2 mu 10 (but mu 15_mu 10_EFFS recovers >10%) – Also 2 mu 10_loose unprescaled (seed from L 1_2 MU 0 but with overlap removal) • MSonly triggers: mu 40_MSonly_barrel (large fake rate in endcaps) • Too high L 1_MU 10 rate at 1. 5· 1033 seed L 1_MU 11 (but -8% efficiency) – Use of RPC high-p. T timing improved by 4% with new alignments and channels fixed – In TGC, from 2 st. to 3 st. : 2. 5 rate reduction w. r. t. 2010 and only 2 -3% inefficiencies ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 16

Egamma Trigger status • Electrons: e 20_medium deployed, but rate is high (reduction by 3) – – Work ongoing for e 20_medium 1 as main electron trigger @ L>1033 Ongoing studies vs. cuts (B-layer-hit, n. Pixel, track-cluster, TRTratio, …) When L 1_EM 14 rate unaffordable (2· 1033), primary will be e 22_medium 1 (seed: L 1_EM 16) L 2 optimization at low p. T for J/ ee triggers (now prescaled, but more purity needed) • Photons: in 2011 g 80_loose used as primary (g 60_loose prescaled to 2 Hz) – Efficiency of g 20_loose as on 2010 data: >99% at ET > 25 Ge. V, and up to 500 Ge. V – Both L 2 and EF rates scale linearly with luminosity in first approximation. ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 17

Pile-up and impact on trigger • Too much out-of-time noise in 2011 affecting nearly all slices: – Already on 2010, more evident when going 75 ns 50 ns BC spacing – Pile-up reweighting needed for more reliable data/MC comparison – Pile-up noise correction is presently running at Tier 0 in offline, first preliminary calorimeter trigger studies look promising so far, and discussion is ongoing in LAr, Tile trigger communities • ETmiss reparametrization of significance being performed • Jets seem not too much affected by noise suppression in central regions (further checks now in forward regions) • Taus have produced new cuts to keep the same efficiency in MC samples (small increase in rates on data) Rejection in tau trigger — Effects on Muon trigger • MC simulation show minor impact on efficiency/resolution but high fake rate • Fake studies ongoing at L 1 • Isolation triggers mostly affected at HLT 18/05/2011 A. Ventura - Trigger in ATLAS 18

Involvement of Italian community in ATLAS Trigger (very quick overview) ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 19

L 1_MU Barrel performance Bologna, Napoli, Roma 1, Roma 3* The group is involved in the monitor, optimization and maintenance of the L 1_MU Barrel trigger (with complete responsibility on the system) 2010 data • Trigger efficiency measurements with both: – inclusive selection on orthogonal streams (using min-bias or jet triggers) – T&P selection on J/ and Z resonances Period 2010 2011 -B 2011 -D MU 10 76% 80% MU 11 68% 70% 72% • Trigger performance looks fine since the beginning of the data taking: timing alignment and pt scale commissioned in details within 2010 – Good match with MC expectations – The optimization of the pt-thresholds is still on-going (tightening the trigger roads to better control the rates) – Visible increase of the trigger efficiency after the winter shutdown due to: h/w fixes, completed timing alignment and RPC detector efficiency optimization (with automatic corrections of HV, following pressure and temperature) – A paper is in preparation on calibration and performance of the system 18/05/2011 A. Ventura - Trigger in ATLAS * M. Biglietti 20

L 1_MU barrel: rates at high luminosity 50 ns minimum 75 ns MU 10 match with standalone muons • Additional L 1 rate observed on the low-pt thresholds, due to “fake muons”: trigger Ro. Is not matched with a reconstructed muon track – Real particles leaving hits in the muon detectors, without a track: very low momentum or coming far from the IP (punch-through, shower muons, bkg from the pipe) – Slow particles (+1, +2 BC): rate increased when LHC bunch spacing changed from 75 to 50 ns – Mainly coming from a restricted region in eta (eta~1) • Filtered-out at L 2 using the track requirements, it’s an important contribution to the L 1 muon bandwidth, that will become critical with increasing luminosity – Work is on-going to understand the origin and find solutions to save low-pt thresholds as long as possible 18/05/2011 A. Ventura - Trigger in ATLAS Delta BC 21

RPC/L 1 offline monitoring Lecce • Offline data quality for L 1 muon trigger (Lecce plus Dresden) includes monitoring of relevant quantities: – – – – Eta-Phi coverage Timing (read-out & trigger) Bunch Crossing ID alignment Trigger distribution vs. trigger sector Trigger condition vs. trigger sector Low/High Pt threshold occupancy Number of working trigger towers High. Pt - run 180401 • Monitoring runs regularly at Tier 0 – For each run produces calibraton constants needed fo MC generation – Constantly updated/maintained with new histograms for DQ shifters Low. Pt - run 180401 https: //atlasdqm. cern. ch/webdisplay/tier 0/1/physics_Muons/run_180481/run/Muon. Detectors/RPC/GLOBAL 18/05/2011 A. Ventura - Trigger in ATLAS 22

RPC studies on calibration stream Roma 2 • LCDS framework already used for the MDTs and TGCs, now includes, developed in Roma 2 , the RPC calibration analysis and runs at the Naples Calibration Center – Starting from RAW calibstream data, tracks are reconstructed; – their intersections on the RPC planes are calculated and stored to a ntuple, together with the RPC PRD – ntuples are custom, very lightweight and contain only the bare minimum info needed for RPC performance measurement; – ntuples are processed by a root macro, which creates txt summaries with all relevant quantities; – summaries from several runs are merged together to achieve the desired statistics. • Operations steered by shifters through a web interface – Using pathena/prun decouples from the specific calibration center and the actual job (ANY analysis can be run in this infrastructure); – grid jobs are sent to our calibration center, but they could be launch wherever the muon calibration stream is being replicated. 18/05/2011 A. Ventura - Trigger in ATLAS 23

L 2 activities: mu. Comb/mu. Iso Roma 1 • 2010: commissioning/improving HLT algorithms in situ – mu. Comb (STACO like standalone-ID tracks combination): • better back-extrapolation of mu. Fast muon at the IP • better parameterization of the resolution functions in curvature, eta and phi – mu. Iso (muon isolation based on ID and/or Calo information) • better treatment of the calorimetric noise and pileup • detailed extrapolation of the muon inside the calorimetric cells, to improve isolation energy resolution (more precise subtraction of the specific ionization of the muon) mu. Comb • 2011: tuning to maximize physics potential – mu. Comb: • tuning of the Ro. I size passed to the ID tracking algorithms • optimize efficiency VS puritytuning the matching window • study in detail effect of pile-up (on-going) – mu. Iso: • speedup algo: pre-caching of calorimetric data requests • tuning for different pile-up conditions mu. Iso correlation with offline – subtraction of the expected pile-up; cuts tuning for up to <8> multiple interactions • specific selection optimized for various physics processes – Z/tau/top physics , low invariant mass isolated di-muons for DY, Exotic physics 18/05/2011 A. Ventura - Trigger in ATLAS 24

Muon HLT performance Lecce • Efficiency measurements (T&P method using Z) at L 2 and EF • Resolution studies with different muon offline references (Muid, Staco) • Studies optimized for SUSY analyses (1 - and 2 -leptons, RPV scenarios) EF_mu 18 − Contribution to 1 -lepton INT note for 2010, now contributing to SUSY 1 -lepton for 2011 HLT 1/p. T combined and standalone resolutions for 2010 Trigger Performance paper - ATL- Endcaps Barrel PERF-2011 -02001, Eur. Phys. J C • Next steps and plans: L 2_mu 40_MSonly barrel, tighter 18/05/2011 − study new signatures solving MU 10 high rate − compute high-p. T dimuon efficiencies − improve fake studies at HLT and investigation on pile-up effects − Data/MC comparison on MC 11 samples A. Ventura - Trigger in ATLAS 25

Muon Trigger efficiencies Napoli, Roma 3* • Efficiency measurements (L 1 and HLT) with: • Orthogonal triggers (Jet. Tau. Etmiss stream) • Tag&Probe method using resonances J/ and Z − Invariant mass ranges: J/ in [2. 86, 3. 34]Ge. V, Z in [80, 100]Ge. V Barrel L 1_MU{0, 15, 20} Endcaps L 1_MU{0, 6, 10} EF_mu 4 data 2010 data 2011 MC J/ψ μμ Recent contributions : - Performance of the ATLAS Trigger System in 2010 (ATLAS-PERF-2011 -02 -001) - A measurement of the ATLAS muon reconstruction and trigger efficiency using J/psi decays (ATLAS-CONF-2011 -021) - Measurement of the differential cross-sections of inclusive, prompt and non-prompt J/psi production in proton-proton collisions at √s = 7 Te. V (ar. Xiv: 1104. 3038) 18/05/2011 A. Ventura - Trigger in ATLAS 26

Dimuon trigger efficiencies • Using the Bayes Theorem we can produce an unbiased measurement of efficiencies ε(EF_mu 4_Jpsimumu) and ε(EF_2 mu 4_Jpsimumu) with muon stream EF_mu 4 with Tag&Probe: p. T – η map 2011 dimuon mass spectrum Tag&Probe can’t be evaluated on data because of PS EF_mu 4_Jpsimumu trigger efficiency: data 2010 vs. MC turn-on curves Muon with higher p. T ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS Muon with lower p. T 27

Muon Trigger Validation • Muon trigger validation team is fully italian Lecce, Roma 1, Roma 3 – Also participating to Rotating Validation Coordination of the whole ATLAS trigger • Continuous validation of muon trigger (L 1/L 2/EF) performance on data and MC samples, production for summer conferences: – https: //twiki. cern. ch/twiki/bin/view/Atlas. Protected/Atlas. Production. Group. MC 10 a – Another campaign @ 50 ns in under way (MC 10 b) – Trigger EDM size in MC AOD is under control ( 50 k. B/event) • Presently validating rel. 16 caches for simulation (MC 10 b re -processing and MC 11 coming) — Getting ready for rel. 17 — Studies w/ and w/o pileup • Automatic validation procedures are constantly run and checked: — E. g. : muon L 2 standalone resolution improvements from 16. 6. 2. 5 (magenta) to 16. 7. 0. 2 (green) in digitization+reconstruction 18/05/2011 A. Ventura - Trigger in ATLAS 28

B-jet Trigger Genova Two classes of triggers under the responsibility of the b-jet trigger group (Genoa plus interests from SLAC and Marseille) • µ-jet triggers with a matching requirement: – a set of triggers with different jet thresholds to cover uniformly the wide jet p. T spectrum – widely used in 2010 and in place in 2011 with small adjustments (better coverage in the very low p. T region) – mainly for b-tagging calibration purposes but also for b-jet cross-section measurement • b-jet triggers: triggers – in active mode in the 2011 menu (not in 2010) – default tagger for this year running: Jet. Prob for both L 2 and EF – three working points defined to give an efficiency of 0. 5, 0. 7 and 0. 9 with respect to a b-tagged offine jet sample – triggers in place for: SUSY with 0 -lepton, VBF H→bb, hadronic top (plus general multi b-jet topology) – first efforts ongoing to estimate trigger efficiency with data-driven methods 18/05/2011 A. Ventura - Trigger in ATLAS 29

Triggers on Long Lived Neutral Particles • Specific triggers developed for H LLNP decays in ID, Calo and MS − − Cosenza, Roma 1 Trackless Jet trigger (l 2 j 30_Trackless_HV): neutral decays in the ID and Ecal Cal Energy Ratio (j 35_a 4 tc_EFFS_L 1 TAU_HV): neutral decays in HCal Muon cluster trigger (2 MUL 1_l 2 j 30_HV): neutral decays in MS 3 mu 6_MSonly and 2 mu 6_MSonly_g 10_loose: lepton jets with displaced vertices Hidden Valley Lepton-jets MSSM production Neutralino V V + LSv. P _ V bb and (MET) • All triggers deployed at P 1 and running also on empty/unpaired to evaluate backgrounds − crucial for LLP decays physics searches (see C. Gemme talk) • Rates ~ 1 Hz for each trigger at 1033 − no prescaling. Muon cluster strigger = 0. 47 nb • Groups involved: Cosenza, Cosenza Rome 1, Rome 1 Seattle ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 30

Conclusions • LHC record luminosity constantly improved (now: 8. 84· 1032) – Smooth ATLAS Trigger operation since start of 2011 – Rate predictions have shown to be very accurate • Large physics program to be explored with new coming data – 2011 trigger menu includes a large variety of interesting signatures – Anyway ATLAS can’t be fully efficient need to make choices and to set priorities for luminosity above 1033 cm-2 s-1 • Performance trigger studies ongoing for all relevant physics channels (and for background control samples) • Italian responsibilities/activities in ATLAS Trigger presently concern mostly L 1, muon + b-jet sectors – Interesting results obtained so far and wide perspectives open – Contributions to other trigger uncovered aspects could be discussed Thanks to the many people contributing to this presentation 18/05/2011 A. Ventura - Trigger in ATLAS 31

Backup and spares ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 32

Other lower rate primary triggers ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 33

Jet Trigger status • On 2010 data, EF Jets were running Cone jet-finding algorithm – Reasonably low trigger rates chains were not run in rejection mode – Inserted new chains with Anti-k. T algorithm, as well as using Topoclusters • Lowest unprescaled threshold at 250 Ge. V (em-scale) – Main primary trigger j 180_a 4 tc_EFFS – Multi-jet triggers with 3 to 6 jets will be available for SM dedicated analyses • Efficiencies computed with orthogonal triggers and bootstrap method ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 34

L 1_MU Barrel rejection/efficiency ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 35

RPC studies on calibration stream ATLAS Italia, Napoli 18/05/2011 A. Ventura - Trigger in ATLAS 36

EF_mu 4 with T&P and EF_mu 4_Jpsimumu Efficiency map Map of errors on efficiency Major: Muon with higher PT 37

Rate Hz Performance H LLNP (L= 26 pb-1) 38 3 mu 6_MSonly strigger Muon cluster = 0. 47 nb strigger = 0. 73 nb 2 mu 6_MSonly_g 10_loose Calo energy ratio strigger = 1. 85 nb strigger = 1. 86 nb Trackless jet strigger = 0. 70 nb Rate vs. Luminosity

Work on trigger infrastructure Activities in Pavia • Maintenance and development of Event Filter infrastructure (EFD, PT) • DAQ monitor services: – Mona. Isa: re-elaborating information in IS and publication of results in IS and OH – DFSummary: status summary of DAQ and of its components • Design and development of Data Flow infrastructure – Unification of L 2, EB and EF on one single farm: • Every single HLT node will execute L 2 processing and (in case of accepted event) also event building and EF processing – Prototype developed and tested at P 1 – Results to be shown during next TDAQ week – In TDIB its approval will be discussed • Main goal is implementation by the 2013 shutdown Italian groups involved in DAQ: Bologna, Genova, Roma 1, Napoli, Pavia. 18/05/2011 A. Ventura - Trigger in ATLAS 39

Italian responsibilities in TDAQ • Members of italian institutes considered here with ATLASwide recognized Trigger responsibilites: GE GE LE NA NA PV RM 1 Coccaro Morettini Ventura della Volpe Izzo Negri Pasqualucci Sidoti Vari ATLAS Italia, Napoli 18/05/2011 b-jet signature coordinator Pixel DAQ rotating trigger validation coordinator L 1 coordinator ROD RPC EF Data. Flow Muon DAQ minbias signature coordinator L 1 mu-barrel coordinator A. Ventura - Trigger in ATLAS 40