1. SiD: Simulation and Reconstruction
SiD Workshop
SLAC
December 15, 2011
Norman Graf
(for the sim/reco group)

2. DBD Goals for 2012
Finalize global system design via detector performance optimization using physics analyses.
Incorporate latest detector R&D results.
Incorporate latest engineering designs.
Improve and understand tracking, PFA and flavor-tagging reconstruction.
Automate & streamline production sim/reco.

3. Detector Development Process
Physics Motivation
Benchmark Reactions
Required
Detector and Accelerator Performance
Detector Design Concept
Parametric Variation with Costs
Evaluate Technical Realizability of Concept
Simulation of Physics and Backgrounds
Research & Development of Necessary Technology
Reconstruction
Engineering Evaluation & Modification of Concept
Analysis: Signal & Background Evaluation
Rational Design?
Benchmark Performance vs Cost

4. Framework Overview APPLICATIONS DATA Event Gen StdHep LCIO Events SLIC
Pythia,
WHIZARD, etc.
StdHep
Event Gen
LCDD
LCIO Events
SLIC
Simulation
LCIO PFOs
Event
Reconstruction
Pandora XML
Tracking
PFA
Compact
LCIO
AIDA,root
Event
Analysis
Flavor-tagging,
Jet-finding, etc.
Publication
Tuple analysis,
Event Display
JAS3,Wired, root, PAW
Tuple Analysis
Visualization

5. Event Generation
Common set of events in stdhep format will be generated for ILD & SiD.
Stored and cataloged on the Grid.
/grid/ilc/prod/ilc/mc-dbd
Will use the B1bTF set of 1TeV ILC machine parameters.
GuineaPig files generated.
lumi-linker files generated.
Signal & background generation soon.
See talk by T. Barklow for details.

6. slic updates Using geant4 9.4.p02 Testing geant4 9.5 LCIO2.0
Comparing new physics lists
Await some guidance from CALICE
SCTG will agree on a common set for the DBD
Testing geant4 9.5
Would like to use 9.5 for DBD production
LCIO2.0
Read spin and color flow from input stdhep event, add to MCParticles.

7. slic updates
If requested, store and save calorimeter hit positions within cells to enable more refined hit calculations.
Will be used to study RPC charge-sharing code.
Can write out each individual calorimeter hit in sensitive layers, but creates large output files, would like to integrate into slic.
Also working to define efficient time history for cal hits.

8. org.lcsim Calorimetry
RPC response will be improved to account for charge sharing across readout pads, electronic noise and inefficiency.
Algorithms will be implemented in reconstruction software, may be migrated to simulation software if deemed appropriate.
GEM and Micromegas responses will also be modelled.
Scintillator response will still be supported, input from CALICE on uniformity, etc. will be used.

9. org.lcsim Tracking I
TrackerHit classes will make use of new LCIO classes which support 1D and 2D hits on planar surfaces.
Track will incorporate new TrackState classes.
Existing collections will be moved.
Improvements in silicon pixel response simulations.
Incorporate non-prompt tracking into the standard reconstruction and used for V finding?
Implementing track finding based on 3D hits.

10. org.lcsim Tracking II
Work ongoing to implement existing Kalman Filter code into standard reconstruction stream.
Extrapolator will be used to calculate TrackStates at Calorimeter face and at the dca to the IP.
Should resolve existing deficiencies in track fits.
Work ongoing to improve overall tracking simulations, lower priority but also useful to other end users.
Effects of field map will be investigated.
Runge-Kutta stepper code being adapted.

11. slicPandora
slicPandora was successfully used to simulate the response of SiD’ (clic_sid_cdr) for the CLiC CDR.
Need to study performance with digital SiD HCal, especially after including improved RPC response.
Number of improvements have been made recently
internal restructuring to provide more flexibility in defining input LCIO collections
accommodation of changes to pandoraPFA itself
Need to understand, characterize and optimize selection of algorithms and their ordering.

12. Vertexing
LCFIVertex has been used for the LOI and CDR to provide secondary vertex identification and jet flavor tagging.
Intend to follow new developments in LCFIVertexPlus and adopt to them as necessary.
not being jet-based offers promise of improved performance in multi-jet environments.
will require some changes in how events are processed.
best to be involved from the beginning.

13. SLAC Scientific Computing Applications
New group formed at SLAC last year
Amalgamation of previous Babar, Fermi, Geant4, Computing Division groups
~25 people providing software application support to experiments at SLAC
Currently no dedicated funding for ILC/SiD/lcsim but proposal being developed to submit to DOE
lcsim is getting some indirect support
WIRED
Support taken over by Dmitry Onoprienko
Effort funded by Fermi (a.k.a GLAST)
JAS/Plotter
Support taken over by Brian Van Klaveren
Funded by Fermi, LSST
Work on new plotter ongoing

14. JAS3 & Wired
Support for spin and color information in stdhep and lcio files has been added to the lcio browser.
Number of improvements added to the Wired event display.
object selection fully synchronized between all panels, tools, and views
keyboard shortcuts
ability to clone views (visibility, selection, cuts settings will remain shared among all cloned views) and to copy views (initial settings for visibility, selection, and cuts are copied from an existing view)
selection is preserved when switching tools/handlers/control panels
WIRED settings are integrated into JAS3 preferences menu
picking options are fully functional
cuts functionality extended.
Manual available at
Plan to implement online web start version (similar to Fermi/GLAST)

15. JAS/Plotter New plotter under development
Already in use by EXO, LSST
JAS3 will be updated to include this plotter
Many improvements over existing plotter
Cleaner design aids extensibility
More plot types
More control over plot style
“publication quality plots”
Cleaner separation of code separates graphics from code
Java 2D
pdf/svg/png etc
HTML5 canvas
opengl/webgl
New feature requests welcome

16. Detector Optimization
Not yet settled on final SiD version to be used for the physics benchmarking.
Detector optimization studies to be undertaken in the next months.
Any specific requests from subdetector systems?
In addition to the baseline, some alternatives could be studied
e.g. all-pixel silicon tracker (SPT), scintillator HCal.

17. Detector Design Baseline: Option 1: Option 2: Option 3: Others?
sid_dbd : LOI geometry with more realistic detector descriptions (~sidloi3).
Option 1:
sid_dbdopt: Detector optimized for 1TeV operations.
Option 2:
sid_dbdspt: Silicon Pixel Tracker option.
Option 3:
sid_dbdsci: Analog scintillator HCal
Others?

18. Detector Implementation
First draft of LOI detector implementation has been available for quite some time.
Need feedback from subdetector groups!
Please see:
What’s missing?
What’s incorrect?

19. Detector Optimization
First iteration of optimization ongoing.
sidloi3 HCal +/- 1
Processed e+e-  WW  qq’, ZZ  qq through full simulation. reconstruction, analysis chain using grid tools.
Use W/Z separation as performance metric.
Useful debugging exercise, thanks to the efforts of Jan, Stephane, Philipp and Jeremy earlier this week, the chain is working.
Results soon.
Need input from detector and analysis groups!

20. The Grid
SiD intends to make full use of Grid Storage Element and computing resources for the DBD detector response simulation and physics benchmarking exercise.
Full tool chain using Dirac has been developed and successfully used to complete the CLiC CDR exercise. Work earlier this week by Jan, Stephane, Jeremy & Philipp has resulted in a working local installation.
Pilot jobs being sent to identify OSG resources.
Need to run full sim/reco chain on small sample of events to identify scope of resources needed.
Will need to identify production manager to oversee benchmark sample generation.

21. DBD Milestones
Learn from and preserve functionality developed for CLiC CDR.
Some amount of global detector optimization.
Adapt tracking code to LCIO2.0 event data model.
Implement Kalman fitting for final track states.
Develop full digital HCal response simulation.
Upgrade to latest PandoraPFA & LCFIPlus.

22. DBD Deliverables Results expected for inclusion in DBD
Full simulation of realistic detector design including support structures.
Overlay of correct admixture of expected beam-related backgrounds.
Full tracker hit digitization and ab initio track finding and fitting.
Digital RPC signal simulation, including cross-talk, noise & inefficiencies.
Full reconstruction using slicPandora & LCFIVertex (LCFIPlus if available)

23. DBD Concerns Issues/concerns for DBD and beyond
Loss of key individuals in reconstruction efforts.
Lack of manpower to conduct detector optimization studies.
Weak connection to subdetector groups
No synergy
Aging software infrastructure.
Delay in generating physics benchmark input events.

24. and beyond…
Techniques developed for ILC and CLiC can also be used for Muon Collider studies.
Some additions to slic and GeomConverter specific to MuC
e.g. tapered endcap calorimeters
Background overlay and timing cut functionality developed and tested at CLiC directly applicable.

25. Muon Collider Shielding
Shielding “nose” fully defined and configurable by specifying zplanes (z,rmin, rmax) for the surface of revolution about the z axis.
Straightforward to study multiple setups.

26. Endcaps
rmax
zmin,
rmin,
opening angle
Layering engine
does the rest.

27. Muon Collider Detector
Initial studies started prior to Telluride meeting.
Will continue in preparation for Snowmass 2013.