GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 1/11 Overview of the Instrument Analysis Workshop Series Eduardo do Couto e Silva September 28, 2004

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 2/11 What is this Workshop Series? It is NOT an –I&T effort –SAS/offline effort But it USES –Data from particle sources recorded during the Integration and Test Phase of the LAT assembly at SLAC –Data Analysis Tools from SAS and I&T It is a LAT Collaboration effort –To support Flight Integration and –To develop ownership of the LAT Instrument

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 3/11 Goals for this Workshop Series Familiarizing LAT Collaborators with the –LAT instrument –Front-End Electronics –Trigger and Data Acquisition –Data Analysis Software –Data taking plans during LAT integration using –Cosmic rays –Van de Graaff photons Create a forum to –exchange knowledge between all subsystems and “hardware and software oriented people” Use simulated and real Data to –exercise reconstruction algorithms (mostly with real data) –exercise the data analysis tools and provide feedback to developers Develop expertise to –uncover and quantify any instrumental effects that could have an impact on the LAT science data –start the work that will evolve into the Science Operations Group of the ISOC –create a core and trained group to participate in the beam tests analysis effort (after instrument delivery) Prepare for Instrument Data Analysis Develop ownership of the LAT instrument

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 4/11 The Workshop Series Instrument Analysis Workshop 1 (June 7-8, 2004) –Kick off meeting –Homogenize the knowledge from people who will do the data analysis –Assign “projects” using Monte Carlo simulated data Instrument Analysis Workshop 2 (September 27, 2004) –Discuss results from projects assigned during Workshop 1 –Discuss results from projects derived from REAL data collected with the Engineering Model 2 (ACD, CAL and TKR) – will start on Oct 1 –Develop a list of instrumental effects that could have an impact on science data analysis –Pretty much “our Readiness Review” for Flight Integration – we are almost there … Instrument Analysis Workshop 3 (TBA 6 weeks before the meeting) –Analysis of real data from the first two towers Instrument Analysis Workshop 4 (Summer, TBR) –Analysis of real data from XX-towers (TBD) “Instrument Analysis Workshop 5” – Collaboration Meeting (Full LAT- TBD) –LAT Data Analysis (and to validate Monte Carlo simulation)

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 5/11 Goals of Workshop 1 - June 7-8, 2004 The focus was on simulated data from the first Two Towers Educate people on behavior and/or performance of –TKR Front-End Electronics –CAL Front-End Electronics –Trigger and Data Acquisition system –Reconstruction software Provide hands-on experience –With SAS/I&T analysis files and tools Use the knowledge acquired during the workshop to –Assign MC projects that will be due during Workshop 2 –Provide input to data taking plans for LAT integration

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 6/11 Goals for Workshop 2 Discuss results from projects assigned during Workshop 1 –Advertise improvements in the I&T/SAS infrastructure to support data analysis Narrow down the work needed for data analysis with real data from the first two towers Bonus from subsystems: –Some highlights of hardware tests prior to delivery to I&T Prepare for Instrument Data Analysis Develop ownership of the LAT instrument

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 7/11 Workshop 1 - Project Assignments 1.Implement dead channels in the tracker for imaging 2.Revisit the spectrum of sea-level cosmic rays 3.Define strategy for implementing Deadtime in MC 4.Validate Energy Scales using CAL EM MC/DATA 5.Compare numbers from alignment procedure to those from metrology at SLAC 6.Calculate the tracking efficiency of each tower using track segments 7.Calculate residuals by comparing CAL and TKR locations 8.Make images of the CAL layers (to expose uniformity of response of the CAL) 9.Make image of TKR layers to identify location of shorted strips and broken wirebonds 10.Implement simulated trigger primitive information into MC 11.How well do we find MIPs (e.g. at several angles, within a tower, across towers)? 12.What is the light output of tracks crossing diodes? 13.What are the effects to the data when zero suppression is applied? 14.What is a “clean” muon definition? 15.Can we find gamma rays and   from showers? Will send a student as part of the long term plan A Truly International Effort

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 8/11 Agenda - Workshop 2 08:00-08:30 - Registration 08:30-08:40 - Overview of LAT Project Schedule - Steve 08:40-09:00 - Introduction to the Workshop 2 – Eduardo 09:00-12:30 - Session 1 - Infrastructure, Trigger and Deadtime - 09:00-09:15 Data Taking Plans for Cosmic rays and VDG photons - Gary/Eduardo - 09:15-09:30 Data Analysis files and info: where to go to get them? - Warren - 09:30-09:45 Overview of distributions for offline reports - Xin - 09:45-10:00 Calibration Trending Database - Xin - 10:00-10:20 Coffee break - 10:20-10:25 Instrument Data Analysis Primer- Eduardo/Lee - 10:25-10:40 Update on Event Display: FRED - Anders/Riccardo - 10:40-11:00 How will we handle SAS code updates during I&T - Richard - 11:00-11:20 Overview of trigger tests and muon telescope - Martin/Su Dong - 11:20-11:30 Trigger Studies - Luis - 11:30-12:00 Instrument Data Analysis Variables (GEM/TEM): Update - Anders - 12:00-12:15 Deadtime in Monte Carlo - Toby 13:30-18:00 - Session 2 - TKR and CAL offline analysis for 2 Towers - 13:30-13:50 Latest and greatest from offline analysis with real TKR data - Michael/Luca? - 13:50-14:20 TKR Recon update - Tracy - 14:20-14:50 TKR Alignment in Gleam - Leon - 14:50-15:05 How Does Calibration Flow Through the New Single CDE CalRecon and Digi? Mark - 15:05-15:15 CAL Calibration Operations during I&T - Eric - 15:15-15:25 CAL Crosstalk Issues and their Implications - Sasha - 15:25-15:40 Crystal Imaging with TKR - Benoit - 15:40-15:50 Light Collection Near the Ends of CDEs - Sasha/Andrey - 15:45-16:00 Coffee break - 16:00-16:20 MIP Selections - Sara/Dario - 16:20-16:40 Data Analysis Bill - 16:40-17:00 List of Data Analysis Tasks: what, who and when - Eduardo/Bill - 17:00-18:00 Discussions and assignment of projects – All 19:00 Laser tag to relieve stress...organized by SAS ( thanks Heather)

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 9/11 Require: Tkr1Gaps > 0 This is was the original idea Image the Tracker Layout Image Example From Bill Atwood

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 10/11 This suggests we require: 1) Tkr1Chisq <.2 2).8 < CalMIPRatio < 1.2 3) TkrNumTracks == 1 Leaves: 4.1 K Events ( ~ 40% Efficiency) This leaves 1 Ineff. Event From Bill Atwood

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 11/11 Selecting MIPs for E2E Trigger & data flow tests 12 layers Si + 3% X 0 of W 4 layers Si + 18% X 0 of W 2 layers Si (no W) layers CsI Each TKR layer constists of 2 Si layers rotated by 90° (X,Y) which contains several thicknesses of W. Thin W Thick W Blank (no W) Each CAL layer consists of 12 crystals and each layer is rotated with respect with the next so that it can provide X and Y measurements µ From Sara Cutini and Dario Gasparrini

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 12/11 24 clusters as expected TKR variables after all cuts 8 clusters as expected 4 clusters as expected About 24 hits About 8 hits About 4 hits Black without cuts Blue with cuts

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 13/11 Tails of distributions CalMIPRatio= 4.4 CalMIPDiff = MeV CalEneSum = MeV Let’s look it with Event Display

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 14/11 “Strange” event This track seems a “clean” MIP, the strange thing is the emission of this electron with high energy (purple track) that end in the CAL, so we can explain tails in previous distributions ~13 GeV µ

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 15/11 Tails in CalTowerGap There is energy in each crystal layer and if we make extrapolation from recon track (white dashed) it seems not hit crystals of one orientation, in fact the two orientation have different length. Gap 0 Gap 4 Gap 3 Gap 5 Gap 1 Gap 2 Gap 6 CalTwrGap = CalTwrGap = 91.6 mm Peaked on 11 mm missed

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 16/11 Correcting for Tower Translations Pink – displacement after translation Blue – NO translation (initial config) Blue (filled) – after reconstruction and alignment were applied From Leon Rochester,Tracy Usher and Hiro Tajima

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 17/11 Correcting for Tower Rotations Pink – displacement after rotation Blue – NO rotation (initial config) Blue (filled) – after reconstruction and alignment were applied

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 18/11 highlow light tapering width (mm) length (mm) CAL Imaging with TKR tracks From Benoit Lott

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 19/11 start of track end of track Lattest Svac tuple Position resolution CAL Imaging with TKR Tracks

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 20/11 Case 1: A given hit produces a trigger request, but its time over threshold is so low that by the time the readout takes place the hit has faded away. (missed hit) Digi Bitword < TEM Bitword Case 2: A trigger request is issued. While waiting for the readout a noise hit takes place (Spurious hit). Digi Bitword > TEM Bitword Case 1 Case 2 * The inconsistent events shown here were simulated with made up rates, actual rates should be different from this. 7/11 Trigger Studies From Luis Reyes

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 21/11 Are all 3-in-a-row combinations equally efficient? Assuming an homogeneous flux and identical layers, one expects every combination to trigger the same number of times: * Entries are not individual events. An event that has a hit in every layer will contribute with an entry in every bin. 9/11 Trigger Studies

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 22/11 Deadtime in MC Set TriggerAlg.deadtime to 0.1 s (like EGRET), require pass dead time trigger condition: Shows the accumulated deadtime (EvtElapsedTime-LiveTime) increasing with event id. Total is 83 s corresponding to 829 triggers. From Toby Burnett

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 23/11 And much more... Nice work on hardware –New and improved CAL Calibrations – CAL/NRL –Discussion on operational settings for CAL – CAL/NRL –Nice study of cosmic rays with TKR stack – TKR/Pisa –Trigger group ramping up very fast – Trigger/SLAC –Studying EM system –proposal to improve trigger information Nice work on Software Infrastructure –Improved Event Display (FRED) – SAS/INFN Udine –Improved CAL digi and Recon - CAL –Improved TkrRecon – SAS/SLAC –GEM/TEM information into analysis files – I&T/SAS –Web based Runs Database and automated run reports – I&T/SLAC Improving Documentation –Instrument Data Analysis Primer – I&T/ISOC/SLAC

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 24/11 Then Eduardo stood up… I spent 1 hour asking for volunteers to work using all possible techniques to persuade people –Invited Peter Michelson for a speech –He threatened to clone me 3 times so that I would bug people more and more –Invited Heather to organize a laser tagging “session” –I offered the opportunity for people to shoot me and Richard to get back at us for the steep learning curve of the analysis software –Used a wood pointer instead of laser pointer –Tried to intimidate people physically This morning a couple of people felt my grief and offered suggestions in how to present things more effectively and so on so forth…If were to listen a little longer they almost convinced me to buil a PMCS schedule ! Discussion of New Projects

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 25/11 Guess what? It worked… Discussion with CAL group will occur on a whole day meeting- Friday –Revisit existing projects –Increase participation from European Collaborators TKR/Pisa/SLAC came with ideas and need help to implement –Need a variable to help out cutting events from edges of silicon to help in efficiency calculations –already got Bill and Leon involved –Planning to give us in next 2 weeks a long cosmic run from the first tower with diagnostics enabled so that we can look at trigger primitives and debugging our infrastrructure Francesco (INFN/Trieste) and Julie (GSFC) –Offered to start looking at zoo events in MC Many people expressed interest in starting playing with the ntuple variables to get familiar with the system –Ohio State (2) –INFN/Perugia (2) – also discuss how to continue Sara’s work –INFN/Bari (2) –ACD group (2) – but not right away Per (KTH/Sweden) plans to support the effort as part a long term effort –Submitted a request for a grant for a student to work with us –Keep fingers crossed Jonathan (Univ Denver) looking for a grant to get a student involved as well

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 26/11 And here comes Peter again So Eduardo do you have enough people now? –Of course not ! –This is not even close to the challenges and excitement for for when we have data –We will keep pestering….

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 27/11 Lessons Learned from Workshop 1 The learning curve to use the SAS software is steep… –Analysis at all levels were performed requiring different skills –“Experts” and “non-experts” provided valuable contributions There were improvements in the infrastructure since workshop 1: –New TKR Reconstruction –Improved Event Display FRED –GEM information in TDS and in the analysis files –SAS Workbook (soon to be available) –Instrument Data Analysis Primer –A draft will be released during this Workshop »Thanks to ISOC for supporting this project –New CAL Calibrations –Additional manpower to the Trigger group –Planning evolved: Overview of End-to-End LAT tests with particle data Develop ownership of the LAT instrument

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 28/11 Web Page

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 29/11 Web Page FRIDAY 8 AM VRVS meetings SLAC Time

GLAST LAT Project Instrument Analysis Workshop September 27, 2004 E. do Couto e Silva 30/11 Special Thanks to Project Management for support –Lowell, Peter and all subsystem managers Everyone who has helped to debug and improve the infrastructure –Thanks SAS for the support and patience with our requests usually due tomorrow… –Sure we want more…we will always do European Collaborators who are providing substantial contribution to keep projects going –Keep the flame alive when away from SLAC… –Come to SLAC if you are feeling lonely… NRL and Pisa for providing the whenever-you-can support during this important construction phase –We appreciate the XX% donations in the weekly meetings –but we need you at a 100% level to understand the instrument! –..and you all came to give talks in both Workshops: thank you ! Everyone working with the “new Kid on the block” : Trigger –We are getting the infrastructure in place –We need to get you some data, MC land is too dry… Bill and Steve for the guidance –Couldn’t do it without you ! Anders, Xin and Warren –Hard-working and dependable: the fuel for this engine –Hope you will be able to get a break to analyze data too…