NIPHAD meeting, Jan. 2006 by Sascha Caron How do we trigger beauty ? The Silicon Track Trigger at D0 The Silicon Track Trigger at D0 and ideas for ATLAS.

Slides:



Advertisements
Similar presentations
Impact parameter studies with early data from ATLAS
Advertisements

Fast b tagging at L2 B tagging meeting XX-XX-04 Sascha Caron (NIKHEF) … using the Silicon Track Trigger (STT) Methods to do a very fast b tagging Some.
Freiburg Seminar, Sept Sascha Caron Finding the Higgs or something else ideas to improve the discovery ideas to improve the discovery potential at.
Current limits (95% C.L.): LEP direct searches m H > GeV Global fit to precision EW data (excludes direct search results) m H < 157 GeV Latest Tevatron.
M.Mevius Open and hidden beauty production in 920 GeV proton –nucleus collisions at HERA-B M.Mevius DESY.
The Silicon Track Trigger (STT) at DØ Beauty 2005 in Assisi, June 2005 Sascha Caron for the DØ collaboration Tag beauty fast …
September 27, 2005FAKT 2005, ViennaSlide 1 B-Tagging and ttH, H → bb Analysis on Fully Simulated Events in the ATLAS Experiment A.H. Wildauer Universität.
Recent Electroweak Results from the Tevatron Weak Interactions and Neutrinos Workshop Delphi, Greece, 6-11 June, 2005 Dhiman Chakraborty Northern Illinois.
Top Turns Ten March 2 nd, Measurement of the Top Quark Mass The Low Bias Template Method using Lepton + jets events Kevin Black, Meenakshi Narain.
Kevin Black Meenakshi Narain Boston University
On the Trail of the Higgs Boson Meenakshi Narain.
July 7, 2008SLAC Annual Program ReviewPage 1 High-level Trigger Algorithm Development Ignacio Aracena for the SLAC ATLAS group.
Silicon Tracking for Forward Electron Identification at CDF David Stuart, UC Santa Barbara Oct 30, 2002 David Stuart, UC Santa Barbara Oct 30, 2002.
In order to acquire the full physics potential of the LHC, the ATLAS electromagnetic calorimeter must be able to efficiently identify photons and electrons.
Chris Barnes, Imperial CollegeWIN 2005 B mixing at DØ B mixing at DØ WIN 2005 Delphi, Greece Chris Barnes, Imperial College.
L3 Filtering: status and plans D  Computing Review Meeting: 9 th May 2002 Terry Wyatt, on behalf of the L3 Algorithms group. For more details of current.
BEACH Conference 2006 Leah Welty Indiana University BEACH /7/06.
Tracking at the ATLAS LVL2 Trigger Athens – HEP2003 Nikos Konstantinidis University College London.
Heavy charged gauge boson, W’, search at Hadron Colliders YuChul Yang (Kyungpook National University) (PPP9, NCU, Taiwan, June 04, 2011) June04, 2011,
Alexander Khanov 25 April 2003 DIS’03, St.Petersburg 1 Recent B Physics results from DØ The B Physics program in D Ø Run II Current analyses – First results.
W properties AT CDF J. E. Garcia INFN Pisa. Outline Corfu Summer Institute Corfu Summer Institute September 10 th 2 1.CDF detector 2.W cross section measurements.
Irakli Chakaberia Final Examination April 28, 2014.
Search for Randall-Sundrum Gravitons with 1 fb -1 of Data Amitabha Das.
Gavril Giurgiu, Carnegie Mellon, FCP Nashville B s Mixing at CDF Frontiers in Contemporary Physics Nashville, May Gavril Giurgiu – for CDF.
August 30, 2006 CAT physics meeting Calibration of b-tagging at Tevatron 1. A Secondary Vertex Tagger 2. Primary and secondary vertex reconstruction 3.
HERA-LHC, CERN Oct Preliminary study of Z+b in ATLAS /1 A preliminary study of Z+b production in ATLAS The D0 measurement of  (Z+b)/  (Z+jet)
Possibility of tan  measurement with in CMS Majid Hashemi CERN, CMS IPM,Tehran,Iran QCD and Hadronic Interactions, March 2005, La Thuile, Italy.
Valeria Perez Reale University of Bern On behalf of the ATLAS Physics and Event Selection Architecture Group 1 ATLAS Physics Workshop Athens, May
Searches for the Standard Model Higgs at the Tevatron presented by Per Jonsson Imperial College London On behalf of the CDF and DØ Collaborations Moriond.
B-Tagging Algorithms for CMS Physics
1 Silke Duensing DØ Analysis Status NIKHEF Annual Scientific Meeting Analysing first D0 data  Real Data with:  Jets  Missing Et  Electrons 
DPF2000, 8/9-12/00 p. 1Richard E. Hughes, The Ohio State UniversityHiggs Searches in Run II at CDF Prospects for Higgs Searches at CDF in Run II DPF2000.
Chunhui Chen, University of Pennsylvania 1 Heavy Flavor Production and Cross Sections at the Tevatron Heavy Flavor Production and Cross Sections at the.
CALOR April Algorithms for the DØ Calorimeter Sophie Trincaz-Duvoid LPNHE – PARIS VI for the DØ collaboration  Calorimeter short description.
Paul Balm - EPS July 17, 2003 Towards CP violation results from DØ Paul Balm, NIKHEF (for the DØ collaboration) EPS meeting July 2003, Aachen This.
Jet Tagging Studies at TeV LC Tomáš Laštovička, University of Oxford Linear Collider Physics/Detector Meeting 14/9/2009 CERN.
The DØ Silicon Track Trigger Wendy Taylor IEEE NSS 2000 Lyon, France October 17, 2000  Introduction  Overview of STT  STT Hardware Design u Motherboard.
Search for High-Mass Resonances in e + e - Jia Liu Madelyne Greene, Lana Muniz, Jane Nachtman Goal for the summer Searching for new particle Z’ --- a massive.
LCWS11 – Tracking Performance at CLIC_ILD/SiD Michael Hauschild - CERN, 27-Sep-2011, page 1 Tracking Performance in CLIC_ILD and CLIC_SiD e + e –  H +
Mike HildrethEPS/Aachen, July B Physics Results from DØ Mike Hildreth Université de Notre Dame du Lac DØ Collaboration for the DØ Collaboration.
Jet Studies at CDF Anwar Ahmad Bhatti The Rockefeller University CDF Collaboration DIS03 St. Petersburg Russia April 24,2003 Inclusive Jet Cross Section.
Susan Burke DØ/University of Arizona DPF 2006 Measurement of the top pair production cross section at DØ using dilepton and lepton + track events Susan.
1 Measurement of the Mass of the Top Quark in Dilepton Channels at DØ Jeff Temple University of Arizona for the DØ collaboration DPF 2006.
4/12/05 -Xiaojian Zhang, 1 UIUC paper review Introduction to Bc Event selection The blind analysis The final result The systematic error.
April 7, 2008 DIS UCL1 Tevatron results Heidi Schellman for the D0 and CDF Collaborations.
Axel Naumann, DØ University of Nijmegen, The Netherlands 04/20/2002 APS April Meeting 2002 Prospects of the Multivariate B Quark Tagger for the Level 2.
DØ Beauty Physics in Run II Rick Jesik Imperial College BEACH 2002 V International Conference on Hyperons, Charm and Beauty Hadrons Vancouver, BC, June.
Moriond QCD March 24, 2003Eric Kajfasz, CPPM/D01 b-production cross-section at the TeVatron Eric Kajfasz, CPPM/D0 for the CDF and D0 collaborations.
B-Tagging Algorithms at the CMS Experiment Gavril Giurgiu (for the CMS Collaboration) Johns Hopkins University DPF-APS Meeting, August 10, 2011 Brown University,
Jessica Levêque Rencontres de Moriond QCD 2006 Page 1 Measurement of Top Quark Properties at the TeVatron Jessica Levêque University of Arizona on behalf.
La Thuile, March, 15 th, 2003 f Makoto Tomoto ( FNAL ) Prospects for Higgs Searches at DØ Makoto Tomoto Fermi National Accelerator Laboratory (For the.
Background Shape Study for the ttH, H  bb Channel Catrin Bernius First year talk 15th June 2007 Background Shape Study for the ttH 0, H 0  bb Channel.
Mark OwenManchester Christmas Meeting Jan Search for h ->  with Muons at D  Mark Owen Manchester HEP Group Meeting January 2006 Outline: –Introduction.
Search for a Standard Model Higgs Boson in the Diphoton Final State at the CDF Detector Karen Bland [ ] Department of Physics,
Viktor Veszpremi Purdue University, CDF Collaboration Tev4LHC Workshop, Oct , Fermilab ZH->vvbb results from CDF.
Marcello Barisonzi First results of AOD analysis on t-channel Single Top DAD 30/05/2005 Page 1 AOD on Single Top Marcello Barisonzi.
Search for Standard Model Higgs in ZH  l + l  bb channel at DØ Shaohua Fu Fermilab For the DØ Collaboration DPF 2006, Oct. 29 – Nov. 3 Honolulu, Hawaii.
Suyong Choi (SKKU) SUSY Standard Model Higgs Searches at DØ Suyong Choi SKKU, Korea for DØ Collaboration.
David Lange Lawrence Livermore National Laboratory
Erik Devetak Oxford University 18/09/2008
Using IP Chi-Square Probability
More technical description:
Venkat Kaushik, Jae Yu University of Texas at Arlington
B Tagging Efficiency and Mistag Rate Measurement in ATLAS
NIKHEF / Universiteit van Amsterdam
The Silicon Track Trigger (STT) at DØ
Tim Scanlon Imperial College, London on behalf of the DØ Collaboration
Susan Burke, University of Arizona
Measurement of b-jet Shapes at CDF
Presentation transcript:

NIPHAD meeting, Jan by Sascha Caron How do we trigger beauty ? The Silicon Track Trigger at D0 The Silicon Track Trigger at D0 and ideas for ATLAS

Why trigger on beauty? – Outline Sascha Caron page 1 The story begins: Motivation for D0 and ATLAS The D0 and ATLAS trigger systems The D0 and ATLAS trigger systems The Silicon Track Trigger at D0 Improving the Silicon Track Trigger B-jet identification algorithms Summary

Tevatron instantaneous luminosity will reach 300E30 (in 1/cm^2*1/sec), a factor 3 increase from the current situation.  Triggering (thus online event selection) increasingly important for optimal performance (Important to note for ATLAS: Events which are lost in the trigger are lost forever!) Dijet background for Higgs  bb or the important calibration process (Z  bb) are in some channels so high that even triggering becomes difficult (at D0 e.g. HZ  vvbb)  Solution : Trigger on b-jets !  Solution : Trigger on b-jets ! Why trigger on beauty? – The Story begins Sascha Caron page 2

NIKHEF interest – My Marie Curie proposal Idea: Having the most efficient b-jet trigger can be “the advantage” for an early Higgs discovery at ATLAS o Test and implement b-jet trigger at D0 o Transform this knowledge (via Freiburg?) to ATLAS Why trigger on beauty? – The Story begins Sascha Caron page 5

B trigger at DØ Sascha Caron page 3 Find b-events early to keep high efficiency at an acceptable rate Find b-events early to keep high efficiency at an acceptable rate Eventspersecond QCD E T >30 GeV dijet production Goals Z->bb, HZ->bbvv, H->bb, etc. Z->bb, HZ->bbvv, H->bb, etc. maybe B physics b-jets E T >30 GeV Z-> b bbar Higgs->b bbar ZH-> bbvv, bH->bbb etc

DØ in Run II The Silicon Track Trigger is based on information of the : Silicon Microstrip Tracker Central Fiber Tracker The Silicon Track Trigger at D0 Sascha Caron page 4

L1 Trigger decision time decision time about 4 μs 2000 Hz 1000Hz 50 Hz 2.5MHz L2 Trigger decision time decision time about 200 μs L3 Trigger decision time decision time about 50 ms D0 Trigger System p p bunch crossing frequency ¯ o Hardware based o tracks made with central fiber tracker, muons, New calorimeter trigger Provides jets, electrons taus (with shower shape) o Hardware/Software o simple jets, electrons, muons, taus o Silicon Microvertex improved tracks (STT)  L2 global processor combines information (e.g. STT tracks for very fast B-id) o Software based o partial event reconstruction (also simple B-id) The Silicon Track Trigger at D0 Sascha Caron page 5

L1 Trigger decision time decision time about 2.5 μs Hz 2000Hz 200Hz40MHz L2 Trigger decision time decision time about 10 ms (50*D0 time) using a farm of 1000 CPUs L3 Trigger decision time decision time about 1s Per event ATLAS Trigger System p p bunch crossing frequency o Hardware based o consists of calorimeter trigger, muon trigger and Central trigger processor Identification of calo. depositions as jets, electrons, taus Muon trigger: muons Position is recorded as Regions of Interest (RoI) o software algorithms (C++ code) o algorithms use portion of data defined in terms of RoI o track fitting using code Pixel and SCT hits as input (or TRT only) o Idea to run a b-tagging algorithm o Software based o (modified) offline algorithms have Full access to event data  event reconstruction Silicon Tracking in the Trigger at ATLAS Sascha Caron page 6

SiliconMicroDetector CentralFiberTracker L2STT L1CTT L2CTT L2Global other L2 pre- processors Detector Level 1 Level 2 Level 1 Detector Data flow chart L2 tracks The Silicon Track Trigger at D0 Sascha Caron page 7

Principal Idea B decay length is mm Impact parameter (2d in x-y plane) B decay products o Silicon Improved Tracks with 2d impact parameter 2d impact parameter o Select events with large impact parameter tracks Interaction point is mean beam spot The Silicon Track Trigger at D0 Sascha Caron page 8

96’ SVT trigger proposal at CDF (first result about 2002) 98’Proposal (by Boston, Stony Brook, FSU, etc.) to build such a trigger at D0 Design algorithms for clustering of Silicon hits, track fitting Design and built trigger cards (FPGAs, DSPs) Write a trigger simulation (same algorithms as online) until 2004: Commissioning (do we understand the data, does the data agree with the simulation, do the trigger tracks agree with the offline (fully reconstructed) tracks) Can we further improve the trigger? How to use the trigger? old NIKHEF interest: top, Z  bb, ATLAS? How do you get such a trigger? Sascha Caron page 9

How can the tracking be improved? o Tracks found at L1 with the Central Fiber Tracker are used to define roads into the Silicon o Silicon hits are clustered o Track is re-fit within the road (IP, χ 2 ) within about 50 µs (IP, χ 2 ) within about 50 µs The Silicon Track Trigger at D0 IP resolution ≈ 50 μm ≈ 50 μm Sascha Caron page 10 Silicon detector Fiber Tracker select event by a cut on IP

Silicon Track Trigger cards Road data to Fiber Road Card - Receives and distributes L1 tracks - Communicates with Trigger framework framework SMT data to Silicon Trigger Cards Cards - Perform clustering and cluster- road matching - Clusters SMT hits + pedestal correction - Axial clusters are matched to the roads Fitting done in Track Fit Card - receives road and axial clusters - convert to physics coordinates via LUT - perform track fit φ(r)= b/r+kr+φ 0 by minimizing chi 2 = Σ i ( φ i - φ(r)/σ i ) 2 - beam spot correction - output tracks to L2CTT The Silicon Track Trigger at D0 Sascha Caron page 11

Commissioning  How does the online clustering work? Compare ADC weighted centroids (online to offline clustering) Sascha Caron page 12 Clustering strips as a function of channel and chip id with 2 treshholds

Clustering agrees between simulation and hardware. Sascha Caron page 13 Commissioning  How does the online clustering work?

At the beginning puzzling results (have to understand offline clustering) Sascha Caron page 14 Commissioning  How does the online clustering work?  Differences all understood : -Bugs in geometry files found -Occupancy cut offline but not online (if more than 25% strips of a SVX chip above treshhold) -Pedestal corrections different (still don’t know why),  After applying this to online clustering 95 % agreement!

Dominant systematic effects on b are beam spot size 30μm and silicon resolution 18μm added into σ 0 ( a=53GeV μm ) STT ansatz for impact parameter significance= b/db calculation: Parameterization of the error db : However the “significance” of the impact parameter heavily depends on the goodness of the fit (we may have taken the wrong hits into the fit) -> Then chi2 gets worse STT track quantities Sascha Caron page 15

Performance studies STT tracks Impact Parameter resolution Correlation to full D0 RECO p T (GeV) Resolution in μ m Purity Including beam spot size 30 μm and 18 μm SMT resolution STT track quantities -- performance Sascha Caron page 16

Compare real STT data with MC  Usual MCs (with Multiple Interaction) to clean to describe the data Sascha Caron page 17 Commissioning  tracks described by MC? Why? - Complete Silicon (small effect) effect) - Multiple Interactions better describable with data describable with data - SMT and track occupancy is luminosity dependent = max(IP/sigma(IP)) LEARN SOMETHING FOR ATLAS!!

Compare real STT data with MC  Usual MCs (with Multiple Interaction) to clean to describe the data Sascha Caron page 18 Commissioning  tracks described by MC? Why? - Complete Silicon (small effect) effect) - Multiple Interactions better describable with data describable with data - SMT and track occupancy is luminosity dependent  Known effect now, will be reduced with new silicon layer and modification of track selection LEARN SOMETHING FOR ATLAS!!

B-Identification methods Idea: Use a very fast L2 b-tagging method to increase performance for e.g. HZ->vvbb, Z->bb events to be run on L2_global processor (time constrain<20 µs)  built 1 discriminating quantity to cut on instead of cutting on maximum impact parameter significance S Compare different b-id methods or invent new ones Sascha Caron page 19

ELIP method Event Lifetime probability (call it ELIP)  Like the old ALEPH method 1) Derive probability p i for each good track to come from the primary vertex Significance distribution is non-gaussian -> p is derived via p (S) =∫ ∞ S pdf(x) dx using a QCD background MC -> can be done very fast via pre-defined lookup tables 2) Derive probability P for each event that all good tracks come from the vertex (or derive prob. That ∏ p i may happen: Time needed to derive P is quite significant (2 loops) Is that needed? No!!  INVENT BETTER ALGORITHMS

Best – algorithm MULM method Significance is heavily dependent on the goodness of the track fit Goodness the track fit given by scaled chi2 (less pt dependent) Idea: Include chi2 information in discriminator by using 2d p(S,chi2) pdfs This degrades tracks with large chi 2 while still using the full information provided by the STT. LEARN SOMETHING FOR ATLAS!! Sascha Caron page 21

One example MC result Z-> b bbar sample: 2 L2 jets with E_T>15 GeV required Signal efficiency much larger for MULM than for all other methods due to the inclusion of chi2 dependence This is a Monte Carlo study (LIKE AT ATLAS, you know)  BUT WE LEARNED THAT WE HAVE TO DO THIS WITH DATA LEARN SOMETHING FOR ATLAS!! CUT ON B-algorithm likelihood CUT ON SIGNIFICANCE

ONLINE ALGORITHM Loop over the 5 ‘good’ tracks with largest IP and derive the product : Derive probability density functions of tracks in B-events : P B and non-B events : P non-B   Store their ratio into a lookup table on the L2 global processor A fast B-id algorithm for Level 2 New Idea: Combine tracks in a fast, multivariate algorithm Probabilityratio P B /P non-B Sascha Caron page 23 P B,i / P non-B,i

A fast B-id algorithm for Level 2 Derive performance of the STT+B-id algorithm with D0 data Data with offlineb-tags Cut method B-id algorithm Sascha Caron page 24 Data without offline b-tag Discriminator of the B-id algorithm Background efficiency Signal efficiency Events First results of a L2 b-tag algorithm

Summary Next step: Build a trigger for inclusive b-jets (take Z  bb as an example) ? Sascha Caron page 25 o Silicon Track Trigger at DZero works o Further improvement by up to a factor 2 with the B-id algorithm with the B-id algorithm  Impact in next trigger strategy for difficult channels

Why is Z  bb and an inclusive signal of any interest? b-jet/light-jet energy scale is dominant error on top mass (0.8 GeV) (error comes from frag., colour flow, semileptonic decay fraction, etc.) Z  bb signal was proposed to reduce this error: error on b-jet/light-jet scale must be < 1% to lead to delta(M_top)<0.8 GeV Sascha Caron page 26

Why is an inclusive signal of any interest? Current strategy for Z  bb analysis: Require muon decay (to have event triggered) + require 2 svt tagged jets (total efficiency is about 0.5 %)  We get about 2 Z  bb events per pb with one B  muonX decay (<0.2 with 2 B  muonX decays) Problem : We have to solve C B  all * P b  all + C B  muonX * P b  muonX = P Z Needs light-to-b-scale correction C b  muon X either from data or MC? Data: We can ask 2 muon tagged b-jets and solve P b  muonX + P b  muonX = P Z How many Z  bb events do we need to measure δC b  muon X  4-8 fb -1 needed to get δC b  muon X ≈ 0.5% … and then we need C B  all ….

Why is Z  bb and an inclusive signal of any interest? We do not want to loose against CDF in delta(M_top)  Other strategy: Measure directly C B  all via an inclusive b-jet event sample (we can apply the same b-tagger on this sample as will be used in top analysis, e.g. NN in the future, will make things simpler) Problem: Can we be trigger Z  bb events without muons in Run2b? Do we get these events through L1 and L2?

L1 condition at 100 E30 A first try: Idea 2 jets with new L1 Jet algorithm (also at ATLAS) Sascha Caron page 29 L1 term L1 efficiency for Z->bb inclusive (v15ttf trigger simulation,Sabine Lammers) CJT_SW2x2(2,10.0,3.2) 41% CJT_SW2x2(2,12.0,3.2) 34.2% at >2500 Hz CJT_SW2x2(2,15.0,3.2) 25.7% at 860 Hz CJT_SW2x2(2,20.0,3.2) 16.2% at 220 Hz CJT_SW2x2(2,25.0,3.2) 7.8 % at 75 Hz CJT_SW2x2(2,25.0,2.5) % at 65 Hz

L2 rate for L1 condition L1: CJT_SW2x2(2,25.0,2.5) L2btagLikelihood cut Sascha Caron page 30

Inclusive b-jet trigger possible for Run2b

Compare with ATLAS Sascha Caron page 31 D0 and ATLAS have 40% efficiency for background reduction of 20  D0 for events!! ATLAS for jets!! (So D0 wins 2:1) (Yes this is a rough estimate and I apologize to the ATLAS L2 btag people) LEARN SOMETHING FOR ATLAS!! ATLAS Very preliminary Taken out from a talk at 24.Okt 2005

How do we trigger beauty? How do we trigger beauty? Within μs a very good construction of Silicon Improved tracks + O(1 μs) construction of a discriminator via a look-up-table b-tag algorithm  Triggering beauty works! LEARN SOMETHING FOR ATLAS!!