Last news on muon time calibration ANTARES-KM3NeT Coll. Meeting Granada, 07/05/2018 Nafis Rezwan Khan Chowdhury Ankur Sharma IFIC INFN – Sezione di Pisa Agustín Sánchez Losa INFN – Sezione di Bari
Since Rabat… • Studies of correction/offset evolution per iteration • Fitting offsets and control plots • Update of the two offsets computed so far (ARCA/ORCA) • Comparison with available offsets • Preliminary findings on MC • Creating a centred documentation of the muon calibration Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 2
Muon time calibration • Determination of detector time constants using atmospheric muon tracks as time reference between detector elements • Only method there where optical beacons are not available… • ...but not a prompt tool: requires many runs for statistics • Recipe: 1) reconstruct tracks excluding the component to calibrate (DOM, DU) 2) collect hit time residuals (HTR) of the tracks with the component to calibrate 3) compare with the rest of the detector HTR to estimate relative corrections 4) include found corrections on the calibration and repeat until reach stability Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 3
DATA MC Program flow KM3NeT_DETID_RUN.root KM3NeT_-DETID_RUN.root JTriggerProcessor / JTriggerReprocessor • Raw DATA method: Event selection: • DATA KM3NeT runs 7 L1 DOM hit • Event selection • Loop on find corrections: JPreSim_HTR • JPreSim_HTR • Offsets per iteration JPreSim+JPreFit track reconstruction + HTR • Final offsets HTR HTR • MC vs DATA method: • DATA & MC KM3NeT runs JFit_HTR • Event selection JFit DATA-MC • Loop on find corrections: HTR comparison • JPreSim_HTR • Jfit_HTR • Offsets per iteration Offsets Offsets • Final offsets Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 4
Studies on evolution per iteration Corrections found per iteration: • Global offset of HTR median at ~1.7-2.0 ns (DU dependent) • Main transitory phase (non-stable region): biggest corrections • Fine tuning phase (stable region): asymptotic small corrections How many iterations required? Evolution of corrections found per iteration • Depends how the offset is computed for ARCA_00005009_00005046 DATA Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 5
Fitting offsets • InterDOM offsets estimation from iteration plots: • A) Average value of last 10 iterations: • Simple to perform and reproduce • Different asymptotic behaviour not into account • B) Fitting to a model: • Complex to define an stable fit • Asymptotic behaviour into account on offset and error estimations • Chosen fit: 𝐸𝑃𝑁𝑡 𝑃𝑔𝑔𝑡𝑓𝑢 𝑗 = 0 per iteration • Simultaneous fit on all the available DOMs plus σ 𝑗 constrain ∙ 𝑓 −𝜇 1 ∙ it + 𝛽 𝐸𝑃𝑁 ∙ 𝑓 −𝜇 2 ∙ it • Fit: 𝑧 𝐸𝑃𝑁 it = 𝑃𝑔𝑔𝑡𝑓𝑢 𝐸𝑃𝑁 + 𝛽 𝐸𝑃𝑁 1 2 • A) – B) differences <0.02 ns on stable DOMs but up to ~0.2ns on difficult cases Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 6
Fitting offsets Asymptotic exponential fitting on ARCA_00005009_00005046 DATA Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 7
#DOM DU001 DU002 1 +8.0 +11.1 2 +8.3 - InterDOM offsets 3 +7.5 +12.2 4 +6.5 -1.4 Computed for ARCA 5 +5.3 +7.7 6 +3.1 +11.0 7 +1.6 +1.6 Code name: ARCA_00005009_00005046 • 8 +1.6 +7.1 38 runs (5009-5046) • 9 -1.3 -2.4 10 -0.9 -5.2 13/12/2016 - 01/01/2017 • 11 -3.3 - DETX: all_DB_PMT_noref.detx • 12 -1.2 - Karel’s detector file 13 -7.5 -11.2 t0: ? • 14 -4.8 -3.0 geo: ? • 15 -6.7 -5.3 w/o MC + JTP • 16 -7.5 -9.3 17 100 iterations (asymptotic fitting) -8.8 -3.1 • 18 - -9.7 Next slide : comparison with Karel’s OFFSETS (ns) (w MC and same detx) Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 8
Offset (ns) +15.0 +15.0 Offset (ns) # # w/o MC w MC DIFF w/o MC w MC DIFF +10.0 +10.0 1 +2.6 1 +8.0 +5.4 +2.8 +11.1 +8.3 InterDOM offset (ns) InterDOM offset (ns) 2 +2.5 2 - +8.3 +5.7 - - +5.0 +5.0 3 +2.3 +7.5 +5.2 3 +2.2 +12.2 +10.0 +0.0 +0.0 4 +1.6 4 +6.5 +4.9 +1.9 -1.4 -3.3 5 +1.3 5 +5.3 +4.0 +1.4 +7.7 +6.3 -5.0 -5.0 6 +0.4 6 +1.2 +3.1 +2.7 +11.0 +9.8 7 -0.0 +1.6 +1.6 7 -4.4 +1.6 +6.0 -10.0 -10.0 8 -0.5 8 +1.6 +2.1 +0.5 +7.1 +6.6 -15.0 -15.0 9 -1.0 9 -1.3 -0.3 +0.2 -2.4 -2.6 1 2 3 4 5 6 7 8 9 101112131415161718 1 2 3 4 5 6 7 8 9 101112131415161718 10 -1.0 10 -0.1 -0.9 +0.2 -5.2 -5.1 ARCA DU001 DOM# ARCA DU002 DOM# 11 -1.2 -3.3 -2.1 11 - - - 12 -1.3 12 -1.2 +0.1 - - - 13 -1.3 13 -7.5 -6.2 • General good agreement -0.6 -11.2 -10.6 14 -1.4 14 -0.9 -4.8 -3.4 -3.0 -2.2 • Two main differences: 15 -1.1 -6.7 -5.5 15 -1.1 -5.3 -4.2 16 -0.8 16 -7.5 -6.7 -1.3 -9.3 -8.0 • At bottom DOM → MC? 17 -1.3 17 -8.8 -7.5 -1.0 -3.1 -2.1 • DU002 DOM07 → Karel’s step fluctuations during 18 - 18 -0.7 - - -9.7 -9.0 iterations? ARCA-DU1 ARCA-DU2 Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 9
#DOM DU002 1 +4.5 2 +4.1 InterDOM offsets 3 +4.3 4 +4.4 Computed for ORCA 5 +3.5 6 +3.3 7 +1.9 Code name: ORCA_00003165_00003192 • 8 +1.4 runs (3165-3192) • 9 +1.1 10 +0.5 04 - 08/12/2017 • 11 -0.4 DETX: KM3NeT_00000029_20170920.detx • 12 -0.7 Same used by Rodrigo for NBs 13 -2.2 t0: ? • 14 -3.5 geo: ? • 15 -4.1 w/o MC + JTR • 16 -5.0 17 100 iterations (average last 10) -6.7 • Next slide : comparison with Rodrigo’s 18 -6.1 (NB and same detx) OFFSETS (ns) Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 10
+15.0 Offset (ns) # +10.0 InterDOM offset (ns) w/o MC NB DIFF +5.0 1 +4.5 +12.7 -8.2 2 +4.1 +10.6 -6.5 +0.0 3 +4.3 +8.4 -4.1 -5.0 4 +4.4 +7.5 -3.1 5 -10.0 +3.5 +5.0 -1.5 6 +3.3 +4.0 -0.7 -15.0 7 +1.9 +3.0 -1.1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 ORCA DU002 DOM# 8 +1.4 +1.4 +0.0 9 +1.1 -0.3 +1.4 A preliminary interpretation of NB (N)-(N+1) Δ t into offsets: 10 +0.5 -1.9 +2.4 11 • NBs seems much more sensitive than Muon -0.4 -3.4 +3.0 12 -0.7 -3.7 +3.0 • Muon method may suffer of some bias on lower DOMs, stronger if MC 13 -2.2 -4.1 +1.9 comparison is used (see ARCA DU1) 14 -3.5 -5.2 +1.7 15 -4.1 -6.9 +2.8 • Both methods suggest the possibility of an incorrect interDOM distance: 16 -5.0 -7.7 +2.7 time/distance slope compatible with a ~3% DU elongation (also on ARCA) 17 -6.7 -10.0 +3.3 • If this is correct: 18 -6.1 -9.3 +3.2 • previous offset plot slope tendencies are explained ORCA-DU2 • …yet ARCA DU2 features are not explained Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 11
Preliminary findings on MC • InterDOM offsets look on ARCA_00005009_00005097 MC: • MC: /hpss/in2p3.fr/group/km3net/mc/KM3NeT_ 00000014/Runs_5009- 5097/JTE/v3.1/KM3NeT_- 00000014_02072016_mu10G_*.km3Hit2_v5r 4.JTERun50*Eff05.root • detx: /afs/in2p3.fr/throng/km3net/detectors/KM 3NeT_-00000014_02072016.detx • No transitory phase (only fine tuning) • Some systematic bias starts to be confirmed… active DOMs dependant? Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 12
Documentation and official offsets • Building a kind of “storefront” of the muon calibration at GIT: • http://git.km3net.de/muon_calibration ./documentation [→ really incomplete at this stage: first step is add all talks] ./code [→ really incomplete at this stage] ./calibrations ./interDOMoffsets [→ ready for the only offsets done so far] ./interlineoffsets [→ for the future] • Easy to communicate to other calibration groups and to the collaboration • Aim to stablish a common language: in order to compare between different calibrations clearly and less prone to • errors: offset format proposal, used runs, methods… • Difficulties in defining unequivocally conditions of a certain calibration: • Assumed geometry → DB table reference? Real -time geometry version + runs? • Assumed t0 initial calibration → t0 table used to build the . detx file? Info at DB? Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 13
Conclusions • Studies on the evolution per iteration: understanding the method • Fitting procedure: more control on offset and error estimations • General agreement between methods: similar things are seen (small differences still to be understood) • Systematic bias evidence in MC: parametrisation required • Incoming plan to study effects with MC: active DOM dependence, etc • Call for coordination and common language-communication standards among time calibration groups Granada, 07/05/2018 Agustín Sánchez Losa – INFN Sezione di Bari | ANTARES-KM3NeT Coll. Meeting 14
Recommend
More recommend
