HCP05 dellAcqua

Preparing for data analysis in ATLAS: 

Preparing for data analysis in ATLAS Andrea Dell’Acqua - CERN PH/SFT <Andrea.Dellacqua@cern.ch> on behalf of the ATLAS collaboration

The ATLAS detector: 

The ATLAS detector Easily the most ambitious (and complex) ever Diameter 25 m Barrel toroid length 26 m End-cap end-wall chamber span 46 m Overall weight 7000 Tons

The road to data taking: 

The road to data taking Two years from the first collisions Activity is becoming hectic… building the detector… building the community… building the software (online and offline) tools… ATLAS takes up the challenge

Building the detector…: 

Building the detector…

Building a community: 

Building a community Only 2 years from data taking and so many things to do… People becoming aware the experiment is coming up and willing to get going at analysis What? How?? Set up a full-scale exercise which looks at the experiment in its initial phase Still, a small-scale exercise when compared to real life…

The Rome Physics Workshop: 

The Rome Physics Workshop 441 registered participants 91 entries (out of about 100 talks), 21 F plus 70 M (preliminary)

The Rome Workshop: 

The Rome Workshop Concentrate on the initial phases of the experiment Priority to e.g. SM physics rather than rare channels Initial layout of the detector Use the latest SW tools First “exposure” (for many) to Distributed production (Grid) Athena Geant4 “Event Data Model” New analysis style/tools Physics results presented at this conference

The Rome aftermath: 

The Rome aftermath The exercise was a terrific success from many respects The “user” community was bootstrapped They like what they used (constructive feedback) They ran the whole chain …and now ask for more and more…. basically all SW tools delivered in a timely fashion and in good shape The meeting set a checkpoint for the computing community: we know now we are going in the right direction… It set also a baseline for future activities

Slide9: 

This is the first successful use of the grid by a large user community, which has however also revealed several shortcomings which need now to be fixed as LHC turn-on is only two years ahead! Very instructive comments from the user feedback have been presented at the Workshop (obviously this was one of the main themes and purposes of the meeting) All this is available on the Web

What is next?: 

What is next? Continuous production and physics analysis Keep the momentum, improve on what we’ve got until now, help the “users” become “experts” Commissioning of the offline computing Aim at having a functional system by mid ‘06 Commissioning of the experiment It’s already happening, now, as we speak Another large-scale physics exercise? Last chance before… Show time…

Offline computing commissioning: 

Offline computing commissioning Major commissioning exercise of all aspects of the offline computing during the first half of 2006 Formerly called “DC3” More a running-in of continuous operation than a stand-alone challenge Main aim of Computing System Commissioning will be to test the software and computing infrastructure that we will need at the beginning of 2007 Calibration and alignment procedures and conditions DB Full trigger chain Tier-0 reconstruction and data distribution Distributed access to the data for analysis At the end (summer 2006) we will have a working and operational system, ready to take data with cosmic rays at increasing rates

The ATLAS Event Data Model: 

The ATLAS Event Data Model RAW: “ByteStream” format, ~1.6 MB/event ESD (Event Summary Data): Full output of reconstruction in object (POOL/ROOT) format: Tracks (and their hits), Calo Clusters, Calo Cells, combined reconstruction objects etc. Nominal size 500 kB/event currently 2.5 times larger: contents and technology under revision, following feedback on the first prototype implementation AOD (Analysis Object Data): Summary of event reconstruction with “physics” (POOL/ROOT) objects: electrons, muons, jets, etc. Nominal size 100 kB/event currently 70% of that: contents and technology under revision, following feedback on the first prototype implementation TAG: Database used to quickly select events in AOD and/or ESD files

Offline SW: Architecture: 

Offline SW: Architecture The architecture of the Athena framework is based on Gaudi: Separation of data from algorithms Separation of transient (in-memory) from persistent (in-file) data Extensive use of abstract interfaces to decouple the various components Backbone of the ATLAS computing system Quite extensively used It scales, it works…

Offline SW: Detector description: 

Offline SW: Detector description The GeoModel detector description system provides us with an application-independent way to describe the geometry In this way Simulation, Reconstruction, Event Display etc. use by definition the same geometry Geometry data are stored in a database with a Hierarchical Versioning System Alignment corrections are applied with reference to a given baseline geometry Time to be even more ambitious!

Offline SW: Simulation: 

Offline SW: Simulation Event generator framework interfaces multiple packages including the Genser distribution provided by LCG-AA Simulation with Geant4 since early 2004 automatic geometry build from GeoModel (~5M volumes) >25M events fully simulated up to now since mid-2004 Digitization tested and tuned with Test Beam Fast simulation also used for preliminary large-statistics (physics) background level studies

Offline SW: Reconstruction: 

Offline SW: Reconstruction Separation of data and algorithms: Tracking code: Calorimetry code: Resource needs (memory and CPU) currently larger than target values Optimization and performance, rather than functionality, will be the focus of developments until detector turn-on

Offline SW: Physics Analysis Tools: 

Offline SW: Physics Analysis Tools The Physics Analysis Tools group develops common utilities for analysis based on the Athena framework classes for selections, sorting, combinations etc. of data objects constituent navigation (e.g. jets to clusters) and back navigation (e.g. AOD to ESD) UserAnalysis package in Athena interactive analysis in Athena analysis in Python interfaces to event displays testing the concept of “Event View”: a coherent list of physics objects that are mutually exclusive any object appears only once in the list of reconstructed objects available for analysis

Slide18: 

Still, all of this is just simulation, right?

The ATLAS Combined Test Beam: 

The ATLAS Combined Test Beam Full “vertical slice” of ATLAS tested on CERN H8 beam line May-November 2004 Geant4 simulated layout of the test-beam set-up For first time, all ATLAS sub-detectors integrated and run together with common DAQ, “final” electronics, slow-control, etc. Gained lot of global operation experience during ~ 6 month run. Common ATLAS software used to analyze the data

Slide20: 

~ 90 million events collected ~ 4.5 TB of data: e,   1  250 GeV  ,  , p up to 350 GeV ~ 30 GeV B-field = 0  1.4 T

Slide21: 

A few very preliminary results

Validating detector simulation at the CTB: 

Validating detector simulation at the CTB ATLAS preliminary

Validating detector simulation at the CTB: 

Validating detector simulation at the CTB

Next stop: detector commissioning: 

Next stop: detector commissioning Phase A System commissioning to ROD level. System commissioning for LVL1 and DAQ Check cable connections. Infrastructure commissioning (refrigerators, water cooling, etc.) Phase C System/Trigger/DAQ combined commissioning Phase D Global commissioning cosmic ray runs, planning for initial physics runs; initial off-line analysis software available, first collisions. Phase B ROD – Local DAQ connections established. Calibration runs on local systems. Skeleton TTC system needs to be available. 1/03 03/04 08/06 11/06 Commissioning with “physics data” starts here

ATLAS is taking data!: 

ATLAS is taking data! ATLAS Tile calorimeter already recording cosmics going through More detectors to follow soon Sub-detector commissioning starting

Slide26: 

Cosmic muons in ATLAS pit in 0.01 s …. From full simulation of ATLAS (including cavern, overburden, surface buildings) + measurements with scintillators in the cavern: ~ 106 events in ~ 3 months of data taking enough for initial detector shake-down (catalog problems, gain operation experience, some alignment/calibration, detector synchronization, …) Through-going muons ~ 25 Hz (hits in ID + top and bottom muon chambers) Pass by origin ~ 0.5 Hz (|z| < 60 cm, R < 20 cm, hits in ID) Useful for ECAL calibration ~ 0.5 Hz (|z| < 30 cm, E cell > 100 MeV, ~ 900 ) Check detector operation with cosmic muons

A “typical” event: 

One track reconstructed in Muon chambers Two tracks reconstructed in Inner Detector Will happen every ~10 s A “typical” event

Slide28: 

Beam-halo Simulation of machine background performed by LHC crew (V. Talanov): -- based on MARS; recent machine optics V 6.4 -- scoring plane at the cavern entrance before ATLAS shielding (z =  23 m from IP)  then particles are transported by ATLAS full simulation Single beam period: beam-halo muons and beam-gas events

Slide29: 

Examples of beam-halo muons in ATLAS A typical snake …

Slide30: 

Beam-gas: -- p(7 TeV) on p(rest) -- vertices uniformly distributed over  23 m -- s(pH, pC, pO, …)  s(pp)×A0.7 (inelastic only) -- vacuum estimate: ~3.10-8 Torr (~1015 mol/m3) Single beam period: beam-halo muons and beam-gas events

Slide31: 

Beam-gas collisions are essentially boosted minimum-bias events  low-pT particles Rate : ~ 2500 interactions/m/s

What more?: 

What more? Detector “as-built” Reproduce the situation as in the pit Calorimeters are “pear-shaped” The LAr barrel in not centered on the beam line A barrel coil is not at its nominal position … Good reproduction of “inert” material Pipes, rails, gangways, elevators, cables and what not… Alignment Field ….

Are we done now?: 

Are we done now? Not quite, yet… EDM evolution (ESD/AOD as per user feedback) Analysis model Distributed production Distributed analysis Tier-0 operations Condition DBs … …but the tunnel is past

Summary: 

Summary The ATLAS detector is coming up, steadily and on schedule. It will be there for the first collisions The ATLAS collaboration is getting prepared for data analysis. All major bits & pieces are falling in place There is still a looong road in front of us, and time is getting tight, but we now believe we can make it Quite an interesting list of things to do even before the first collisions