Optical Reconstruction of Proton Decay Events " " - - PowerPoint PPT Presentation

Optical Reconstruction of Proton Decay Events

Kevin R. Wood kevin.wood@stonybrook.edu June 22, 2016

𝑞" → 𝐿"𝜉̅

• Characterization of photon detectors + reconstruction performance.

Revisiting studies done last year* with 200MeV energy deposition and simulated

• information. Now we take advantage of reconstructed proton decay events.
• Attenuation length
• Reconstruction efficiency (flash finding)
• And, additionally, timing resolution
• PDK photon reconstructed signals
• Towards incorporating PD information into PDK background rejection
• Effect of radiological activity

* http://lbne2-docdb.fnal.gov:8080/cgi-bin/RetrieveFile?docid=10896&filename=2015-05-19- Himmel-PDSim-Review.pdf&version=6

6/22/16 2 Optical Reconstruction of PDK Events

Overview

• Gabriel Santucci has generated, simulated, and reconstructed a sample of 10,000

“golden channel” proton decay events

• /pnfs/dune/persistent/users/gsantucc/correct_flat_dist/reco/8011111_*/p2K+nuba

r_2_g4_detsim_reco.root

• Uniformly distributed inside

the 1x2x6 geometry

• Reconstructed using standard

configurations (v05_13_00)

• standard_reco_dune10kt_1x2x6.fcl

6/22/16 3 Optical Reconstruction of PDK Events

𝑞" → 𝐿"𝜉̅ sample

𝑞" → 𝐿"𝜉̅ sample

𝑞" → 𝐿"𝜉' ( 𝐿" → 𝜈"𝜉* 𝜈" → 𝑓"𝜉'𝜉*

𝜐-~1012𝑡 𝜐*~1014s 𝐶𝑆 = 63.5%

6/22/16 Optical Reconstruction of PDK Events 4

𝜈" 𝐿" 𝑓"

0.4 GeV

Attenuation

(A. Himmel)

6/22/16 5 Optical Reconstruction of PDK Events

• LHS incorporates APA edge effects and reconstruction
• Reconstructed information from PDK suggest an attenuation length of ~114 cm while the

photon libraries give ~136cm

Energy deposited into the central region of the APA (yz plane); uses photon library information

• How often do we reconstruct a flash given a PDK event? (5PE threshold)
• Covers everything at >97.5% efficiency (early + late light). >99% until last ~13cm.
• Appreciable portion of Kaons deposit <200MeV, which affects our flash finding.

Optical Reconstruction Efficiency

(A. Himmel) PDK 200 MeV

6/22/16 6 Optical Reconstruction of PDK Events

Optical Reconstruction Efficiency

• Threshold for flash finding on the previous slide was set to 5 PE.
• Radiological backgrounds (and noise levels) dictate where this threshold

should be set.

(From 10K PDK) 2 PE threshold finds everything

6/22/16 Optical Reconstruction of PDK Events 7 39Ar

The effect that the presence of 39Ar has on efficiency/purity will be studied in more detail.

Timing Resolution

• Time resolution O(10 ns)
• TDC DAQ sampling rate of 500 ns would make

this negligible

• Important for non-beam events
• PD gives us T0 and therefore X coordinate

(distance from APA)

• (ttrue is the time associated with the Kaons)

6/22/16 Optical Reconstruction of PDK Events 8

PE Yield vs. Kaon Momentum

6/22/16 9 Optical Reconstruction of PDK Events

• Characteristic photon yield from PDK ranges from a few hundred PEs to a few thousand PEs
• Can we use PD information to help separate PDK from backgrounds (e.g. atmospheric, cosmogenic)?

Summary/Moving Forward

• We’ve improved upon some initial estimations of the PD performance with a

sample of 𝑞" → 𝐿"𝜉̅ events

• Non-beam events drive the PD requirements
• Found reasonable agreement, but some expected deviations from the cruder

study (200 MeV, no reconstructed information)

• ~114cm attenuation length
• >99% reconstruction efficiency in all but the last ~15 cm furthest from the APA

where it dropped to ~97.5%

• Time resolution O(10 ns)
• Begin incorporating radiologicals and cosmogenics into the study
• Also start thinking about incorporating PD information into PDK signal/background

separation (see A. Higuera’s work)

6/22/16 Optical Reconstruction of PDK Events 10

Backup

(new) vertex distribution If you have been using the PDK sample G. Santucci has produced, make sure you are using the newly (in the past week or two) generated sample with these vertex distributions. See slide 3 for location. Old vertex distribution on next slide.

Backup

(old) vertex distribution

• This is from the sample that lives in /pnfs/dune/persistent/users/gsantucc/bugfix/*
• Gaussian… why?
• Buildup on detector edge: most of these tracks will exit the detector.
• If you happen to be using it... don’t. Should be deleted soon