Ice Cube String 21 reconstruction LLH reconstruction algorithm
Ice. Cube: String 21 reconstruction • LLH reconstruction algorithm • Reconstruction of digital waveforms • Muon data reconstruction • Time calibration verification with muons • Combined energy/positional reconstruction • PMT saturation / OM sensitivity • Flasher/muon energy estimate • Timing/geometry verification with flashers Dmitry Chirkin, LBNL Presented by Spencer Klein
Reconstruction in fat-reader contains a plug-in reconstruction module, which: • uses convoluted pandel description • uses multi-media propagation coefficients • relies on the Kurt’s 6 -parameter depth-dependent ice model • has Klaus’s stability of the solution • parameterization is possible for bulk ice • reconstructs both tracks and showers/flashers • calculates an energy estimate • also reconstructs Ice. Top showers • feature extracts waveforms using fast Bayesian unfolding • corrects the charge due to PMT saturation • accounts for the PMT surface acceptance • combines energy with positional/track minimization
LLH Reconstruction
Reconstruction of the simulated data
Root-fit waveform pulse reconstruction
Bayesian waveform unfolding • fast waveform feature extraction: 2 -3 ms per every WF (cf. 30 seconds before) • why not invert against the tabulated smearing function • need to emphasize SPE signal while controlling oscillations of the solution due to noise • Bayesian or regularized unfolding does just that
Bayesian waveform unfolding If a fitted pulse does not start on the boundary, then it is approximated by a superposition of 2 pulses. The weighted average of these pulses gives the estimate of the leading edge. Simple and complicated waveforms are reconstructed with the same amount of effort
Data reconstruction
Comparison with the simulated data
Muon time calibration verification • reconstruct muon tracks without DOM X • plot the time residual for DOM X for nearby reconstructed tracks • if scattering length is longer than the distance cut (10 m) the most likely residual should be 0, otherwise residual will show delay increasing with the amount of scattering.
Energy reconstruction From Gary’s talk: usual hit positional/timing likelihood energy density terms From Chrisopher W. reconstruction paper: Therefore, w=1
Flasher/cascade energy reconstruction The energy estimate m is • constructed according to the Rodin’s Monin Similar treatment for muons formula, with average propagation length obtained from average absorbtion and scattering. These are calculated as during the positional reconstruction, using George/Mathieu prescription based on Kurt’s ice model
PMT saturation From Bai’s DOM test report As measured by Chiba group at 1. 17. 107 Qcorr=Q/(1+Q/Qsat) Qsat=7500 (gain/107)-1. 24 may require new calibration type? Measured between 700 and 1750 V
PMT saturation in flasher data DOM 30 flashing at 127 FFF 20 ns DOMs 29 and 28 show approx. 4600 and 3070 PEs After the correction for saturation DOMs 29 and 28 turn out to receive 11700 and 5100 PEs
OM angular sensitivity From Ped’s thesis, at the moment as parameterized for an AMANDA OM
saturation sensitivity PMT saturation and OM sensitivity
Combined positional/energy reconstruction Improves positional reconstruction by constraining the energy observable: • Systematic position offset is less than 5 meters in all cases • better parameters of Rodin-Monin formula will constrain energy observable even further
DOM-to-DOM variation Fixing position according to the geometry file, and performing only the energy reconstruction Large variation is likely due to ice layering, not entirely inconsistent with a constant. For 03 F/127 one obtains 10^(7. 53) ph. area [m 2].
PMT effective area Measurement at Chiba Chris Wendt’s estimate: 8. 109 +- 20 -50% photons (~56 Te. V) per flasherboard at FFF/127(20 ns) Average quantum efficiency = 0. 165 PMT area = 492. 10 cm 2 81 cm 2 effective area Nph(03 F/127) = 4. 2. 109 photons (for 6 LEDs) At FFF/127(20 ns): 8. 4. 109 photons Cascade: 1. 37 105 photons/Ge. V Energy = 61 Te. V
Muon energy reconstruction Energy density llh term constrains muon-to-string distance, 90% of muons pass within 24 meters of Based on: the string. Still, for MC 90%-3 of muons pass within. -1 Area Nc [m] = 32440 [m ] (1. 22+1. 36. 10 E/[Ge. V]). 81 cm 2 34 meters. A pull toward the string still exists. A better distance estimate results in better resolution in energy of the muon Average measured energy vs. surface energy of simulated muons
Flasher timing information Nearby or in clear ice followsverified expectation from geometry Flashing DOM X we can measure arrival time of the first photon at DOMs above and below. Those that form sharp distributions can be used for timing jitter measurement (rms of the ditribution) and geometry verification (mean).
Conclusions • llh algorithm results in estimates position and energy • 3 methods of waveform feature extraction are implemented • muon and flasher positional reconstruction are satisfactory • muon and flasher energy reconstruction work, but need improvement, based on better pdf and OM sensitivity • timing and geometry are verified with muon and flasher data • an icetray reconstruction module I 3 llh. Reco exists (to be released in ~1 week by Jon Aytac) • Work is done on multiple muon selection and reconstruction
- Slides: 22