1 / 38

Jets & Missing E T

Jets & Missing E T. G. Dissertori N. Varelas. Both for Jets and MET : Present status ongoing validation activities Plans for the near term - work to be done Group structure News Summary. CMS Italia, Napoli Feb 13, 2007. Detector -> Jets & MET. Calibration constants

jake
Télécharger la présentation

Jets & Missing E T

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. Jets & Missing ET G. DissertoriN. Varelas • Both for Jets and MET : • Present status • ongoing validation activities • Plans for the near term - work to be done • Group structure • News • Summary CMS Italia, Napoli Feb 13, 2007

  2. Detector -> Jets & MET Calibration constants used up to this point are a responsibility of DPG ECAL Calibration to p response DPG JetMET Group Note : to be also considered in the future : Different input to Jet Reco and MET Reco, such as Particle Flow Objects, charged tracks, …

  3. JETs

  4. NIC = Noise In Cone (GeV) JEL = Jet Energy Loss (GeV) T scheme was used in most ORCA JetMET studies: ET>0.5 GeV & E>0.8 GeV B scheme used by default in CMSSW JetMET Reco An additional tower energy constrain of ET>0.5 GeV (E>1 GeV) is used in CMSSW Jet cone (kT) clustering PTDR: Jets • Basic/Calo/Gen Jet Algorithms: • Iterative Cone (IC) • Mid Point Cone (MCone) • Inclusive kT • Calo Inputs: • Cells contribute to tower energy if they pass energy thresholds ORCA: ET recombination scheme CMSSW: E-scheme

  5. Status: Jets • Jet Workshop (JW): Dec. 13-14, 2006 • Jet finding at all stages: Trigger - Reconstruction - Simulation • Goal: • Get a complete picture regarding: • the physics needs/issues with jets • the jet algorithms currently implemented at all levels, along with their performance • the status/prospects of calorimeter/jet calibration • 20 presentations • Agenda: http://indico.cern.ch/conferenceDisplay.py?confId=9316 • Sub-Sessions • Physics with Jets (SM, BSM, HI), summary of TeV4LHC, TEVATRON/HERA experience, Pflow, Jet Algos at CMS, Simulation/Generator, Calibration/Performance of Calorimeter, Triggering on Jets

  6. JW Summary: Jet Physics • Physics Needs: • Lowest ET thresholds used in most areas: ~ 30 GeV • Uncorrected: ~20 GeV J.d’Hondt (13.12.06)

  7. Les houches hep-ph/0604120 kT D-parameter JW Summary: Jet Algorithms (1) • Implemented in CMSSW: • Iterative Cone, Midpoint, kT • Run on generic collection of 4-momenta (candidates) • kT algorithm: • Default (size parameter) D=1; apparently not optimal;study of smaller values requested (ATLAS: D=0.6) • Current CMSSW implementationknown to be buggy • About to be implementedas external package

  8. JW Summary: Jet Algorithms (2) • Fast version of kT (FASTJET) : • indeed very fast (à la Iterative Cone) • Very interesting, of course also for Heavy Ion physics • Question raised: implement it even at HLT level? • FASTJET: new pile-up subtraction scheme proposed, to be looked at RelVal/2006/9/26/RelVal101BJets50-120/GEN-SIM-DIGI-RECO/ A. Heister, D. Kcira, A. Oehler

  9. JW Summary: Jet Algorithms (3) • CONE algorithms (“the devil lies in the details”): • eg. Investigate issue of unclustered energy/dark towers as seen at the TEVATRON • “2nd pass” option considered • “Search Cone” option (currently implemented and used in someORCA studies) NOT recommended by TeV4LHC • Algorithms in CMS and ATLAS: • Is a “sharing” needed? First let’s understand if it is possible • Recommendations from TEVATRON : 1)Use the same algorithms in the two experiments!2) Use both CONE and kT algorithms! M. Zielinski (13.12.06)

  10. P.Janot (13.12.06) JW Summary: Particle Flow • Different approach: • use objects/particles identified by particle flow algo and individually calibrated as input to jet algorithms • In principle should need much smaller jet correction factor • However, not as “straightforward” as the CaloTower approach • Close communication needed with HCAL/ECAL tower approach, understand the interplay/needs/impact

  11. JW Summary: Calibration/JES (1) • Tevatron experience / recommendation • Understanding of JES needs TIME and MANPOWER: start now to build a strong group (“dream team”) • Have the tools ready “now”…. • ECAL • Software tools for calibration workflow and for mis-calibration studies are in place • Currently discussing on the final calibration strategy • No detailed study yet on interplay between pure e.m. and pion/jet calibration: to be done

  12. JW Summary: Calibration/JES (2) • HCAL • DPG-calibration task: Produce all constants needed to provide the energy in each tower • Example test beam result: Longitudinal Shower Profile for  in HB prefers G4 LHEP over QGSP used at the moment… • Important task ahead: Pion response calibrated by taking into account the e/h response in ECAL & HCAL • With additional corrections for non-linear effects : Hope to get an energy resolution in HCAL of ~ 80% for stochastic term & 8% for constant term using these methods M.Velasco (14.12.06)

  13. From D0 JW Summary: Calibration/JES (3) • A complicated multi-step approach • Have to investigate a large set of different physics channels to obtain angular uniformity, relative and absolute energy scale • Example: di-jet balancing: Detailed study needed of systematic effects • Example: photon-jet: two calibration techniques investigated • At the moment also: MC-based corrections, tools in place and applied to large productions (for CSA06 and for 1.2.0, no PU yet)

  14. JW Summary: Calibration/JES (4) A.Bhatti (14.12.06) A lot of space for many tomake important contributions

  15. Recent Validation Studies: • MCJet energy corrections • R. Harris (Fermilab), S. Petrushanko (Moscow) • Jet response and resolutions • J. Widawsky, M. Zielinski (Rochester) • Jet efficiencies • L. Apanasevich (UIC) et al. • Jet corrections • Dijet balance: A. Bhatti (Rockefeller) et al. • Photon+jet: O. Kodolova (Moscow) et al. • MET resolutions & total event ET • R. Cavanaugh, B. Scurlock (Florida)

  16. PTDR(Vol.1) Validation Plots: Jets (1) P.Janot SDPV (20.12.06) Fig.s 11.3/4/6/7/8/9: jet reco efficiency, energy response, energy/angular resolution vs pT and 

  17. Barrel Endcap Forward MCJet Corrections • Corrections to particle level based on QCD Dijet Monte Carlo • Methodology developed by Moscow group for ORCA; ported to CMSSW, now LPC and Moscow groups working on it • Corrections available for Icone5 and Mcone5 in 1_2_0 & CSA06 • Accurate within a few %, except for low Gen ET < 30 GeV • At low ET, poor jet resolution makes large corrections difficult • Response = CaloJet ET/GenJet ET • Barrel: 1_2_0 within a few percent of ORCA • Endcap: Miscalibration of HE in CSA06 data fixed in 1_2_0 • Forward: HF Shower Library in CMSSW different than in ORCA

  18. Raw Icone5 Energy Response:1_2_0 vs ORCA ORCA 1_2_0 • Reasonable comparison in HB and HE • A big difference in HF, understood due to problem in HF shower library

  19. Corrected Icone5 Response vs h and ET • Closure tests: corrections work within a few %, except for low Gen ET < 30 GeV (due to poor jet resolution at low ET)

  20. Corrected Icone5 ET Resolutions:1_2_0 vs ORCA 1_2_0 ORCA • Corrected resolutions are similar between 1_2_0 and ORCA; problem in angular resolution understood by now

  21. Plans: Jet Algorithms (1) • Validation/Performance Studies • To be completed by end of this month • Develop a default Jet ID criteria • Re-examine default input thresholds for all algorithms • Flexibility (different input configurations for each algorithm) vs. simplicity (a common set of thresholds for all algorithms, even MET) • Optimize reconstruction efficiency vs. fake rate • Working together with the ECAL/HCAL DPGs • Iterative and Midpoint Cone Algorithms • Re-examine default parameter space • overlap fraction • cone area fraction (currently is set to =1) • cone radius • un-clustered energy – 2nd pass jets

  22. Plans: Jet Algorithms (2) • kT Algorithm • Continue supporting standard kT algorithm with D=1 • Study in detail FastJet (now external package in CMSSW) • Support both D=0.6 and D=1 • Work with the physics groups towards an optimized default D-value • Studies of UE and PU energy subtraction scheme as proposed by Cacciari and Salam • Software Tools • Attilio Santocchia is working on jet matching package • Detector - Particle - Parton Jets, Flavour assignment • Study jet triggering scenarios at HLT • Some requests raised to consider even kT at the HLT

  23. Plans: Jet Algorithms (3) • Jets from other input collections • Jets from tracks only • performance, resolution, corrections, adapted to which analyses, topologies? • Jets from Particle Flow objects (same questions as above) • Jets at the lowest ET range • Very difficult area, but important in many analyses, eg. with jet vetos • Jet corrections • Study impact of electromagnet. energy fraction of jets • Impact/Contribution of ECAL to be studied in much more detail • Develop flexible and data driven scheme (see also next slide) • which is also able to respond to quickly varying beam conditions (PU)

  24. Plans: Jet Energy Calibration • Need to establish a data-driven calibration method for all jet algorithms • Rely on experience from Tevatron/HERA experiments • Many tools in place from earlier studies • …but there is much more to be done at all levels • Establish release goals and review procedure • Long list of correction factors: • Offset: Pile-up/noise/underlying event • Calorimeter energy response • Uniformity across eta, phi • Showering effects • Jet overlap effects • Hadronization effects for correcting back to parton jets • Heavy-flavor jets • … • A lot of combinations: • Most corrections depend on jet algorithm type, algorithm parameter set, CaloTower vs Particle inputs, etc. • Bottom line: Need to intensify and prioritize the effort now!

  25. Missing ET Energia mancante…

  26. PTDR Vol. 1: MET n towers • Missing Transverse Energy • Low luminosity Pileup included • <MET> from QCD • Stochastic term  123%√ET •  1700 GeV ET 700 GeV PT dijets   50 GeV observed MET • MET  Resolution • Low MET : approaches Jet size • High MET : approaches calo cell size R.Cavanaugh (2.11.06) QCD ORCA

  27. PTDR Vol. 2: MET • Need • good communication analysis  MET study group • Need • to impose, perhaps to revisit the JetMET recommendations for MET, systematics etc. (partially done for the PTDR) • Need • to study MET corrections schemes (not all used/tried so far) • Need • to prepare for MET filter for real data, protect against detector and accelerator background • Need • to get rid of too low MET cut analyses (some sneaked into the PTDR) • Need • a systematic comparison of FAMOS/ORCA/CMSSW • Energy flow techniques should help to improve the MET resolution Albert de Roeck

  28. MET at DØ NOT SUSY! MET Workshop • MET Days: Sep. 25-26, 2006 • Organized by R. Cavanaugh and M. Spiropulu • Focused on Commissioning • Presentations on: • Tevatron experience • Physics input • Trigger • Calibration • ORCA validation studies • DQM • Commissioning • CMSSW “first look” studies • Agenda: http://indico.cern.ch/conferenceDisplay.py?confId=6002 • A long “ToDo” list…

  29. Ongoing MET validation studies • Not a lot of people working on it for the moment • Effort just started, here early results using CSA06 samples • 1_2_0 samples with and w/o PU under investigation • Total event ET (SumET) and METx distributions • METx resolution vs SumET • Both SumET and MET are global (calorimeter) variables • Especially sensitive to differences in cell thresholds and pileup conditions between ORCA and CMSSW samples

  30. Fig. 11.10, min-bias w/out PU QCD: 0 < pt-hat < 15 w/out PU, Scheme-B thresholds Min-bias w/out PU, 0.5 GeV Thresholds on HCAL cells <SET>~22 GeV <SET>~190 GeV CSA06 PTDRv1 PTDRv1 sX~2 GeV sX~6 GeV CSA06

  31. Fig. 11.11, QCD: 0 < pt-hat < 15 w/PU QCD: 0 < pt-hat < 15 w/out PU, Scheme-B thresholds PTDRv1 QCD: 0 < pt-hat < 15 w/PU, 0.5 GeV Thresholds on HCAL cells <SET>~22 GeV <SET>~396 GeV CSA06 PTDRv1 sX~10 GeV sX~2 GeV CSA06

  32. Plans: MET (1) • Validation of MET • Low level (together with the HCAL/ECAL DPG): • Optimization of calorimeter energy thresholds • Online/offline “cleanup filters” • Detector and data quality monitoring • High level • Comparison to ORCA (coordinated with the SDPV group) • Physics performance • MET @ L1/HLT • Corrections/Calibration at the trigger level • Thresholds, prescales, rates… • Plans for 900 GeV running

  33. Plans: MET (2) • Look at other approaches to MET • Construct MET from recoil of high-pT objects • Using cone clustering • Using kT clustering (should not leave any unclustered energy) • Study for which analyses/topologies this might be advantageous • MET Calibration • Type-1 from corrections to jets • A lot of knowledge gained in the jets area to be transported to MET • Type-2 from corrections to CaloTowers (pile-up, UE, p/e, …) • Muon/electron/photon corrections • Interesting/difficult : very high-pt muons • Residual corrections (min-bias, QCD di-jets, Z/W+jets, …) • Understand impact and handling of varying beam conditions • Look at METperp, MET significance • MET Commissioning for early physics

  34. Plans for MET (3) Note:We need help in ALL efforts of JetMET, in particular for MET !Your contributions are most welcome here.A challenging and important work to be done!

  35. JetMET Group N. Varelas G. Dissertori Group Structure Missing ET Jet Energy Scale Jet Algorithms Group Representatives Simulation/MC Production Tulika Bose Onl Selection Lenny Apanasevich HCAL DPG Olga Kodolova Michal Szleper Software Robert Harris (Jet) Fedor Ratnikov (Jet) Bobby Scurlock (MET) Physics Tools Attilio Santocchia ECAL DPG TBA Particle Flow/t Joanna Weng

  36. News/Summary • MC pass with CMSSW 1.2.0 getting close to completion • QCD, g+jet, W+jets: w/ and w/o PU • https://uimon.cern.ch/twiki/bin/view/CMS/HighLevelTrigger06ExerciseMCSamples • Will provide a reference sample for near term JetMET activities • Jet energy scale corrections for CSA06 and 1.2.0 (w/o PU) are available • Understanding of the Jet/MET performance is steadily improving in CMSSW • Group priorities for the next couple of months: • Complete the Jet/MET validation coordinated by SDPV group • Launch of the sub-groups and underlying activities • JetMET Workshop in the Spring (probably mid April) • Focus will be on “Calibration Path for Jet/MET”

  37. News • We are working on a JetMET “project plan” for 2007 • Identifying tasks • …and the related peoplepower • If we haven’t contacted you already in person, then please: • Indicate to us (email, phone) if you have interest in JetMET activities • Your physics area of interest & tasks that you would like to work on • CMS Week (Feb 26 – March 2) • There will be a JetMET session – please plan to attend • Bi-weekly group meetings will start on March 8 • Note : tomorrow, 16h, we will have a JetMET meeting! • Subscribe to the CMS JetMET HyperNews list • Group’s twiki page under construction

  38. At LHC almost all the time we will see Jets and MET! … and sometimes we will see leptons and photons… CMS-Italia :siete “invitatissimi” a partecipare al lavoro in quest’aria importantissima C’e’ un bel mucchio da fare…. Grazie per l’invito Summary

More Related