80 likes | 156 Views
SPICE Mie: update from last week. Dmitry Chirkin, UW Madison. Q&A: oversized DOM treatment. This is a crucial optimization: For ice model fitting factor x16 is used, takes 7 - 30 days to fit the ice. If using x1, takes 5 – 21 years to fit, next ice model by year 2015-2031.
E N D
SPICE Mie: update from last week Dmitry Chirkin, UW Madison
Q&A: oversized DOM treatment This is a crucial optimization: For ice model fitting factor x16 is used, takes 7 - 30 days to fit the ice. If using x1, takes 5 – 21 years to fit, next ice model by year 2015-2031. The oversize model was chosen carefully to produce the best possible agreement with the nominal x1 case (see next slide). Some bias is unavoidable since DOMs occupy larger space: x1: diameter of 33 cm x5: 1.65 m x16: 5.3 m This could be the reason for ~5-10% variation around 1 in the DOM occupancy ratio to data nominal DOM oversized DOM oversized ~ 5 times photon
Timing of oversized DOM MC Flashing 63-50 63-49 63-48 63-51 63-52 64-50 64-48 64-52 xR=1 default 1 ns xR=1 default do not track back to detected DOM do not track after detection no ovesize delta correction! do not check causality del=(sqrtf(b*b+(1/(e.zR*e.zR-1)*c)-D)*e.zR-h del=e.R-OMR 10 ns
toff vs. fSAM • Reminder: scattering function = HG*(1-fSAM)+SAM*fSAM • both HG and SAM taken at the same g=<cos q> • fSAM>0 appears to shift the front of the distributions (see next slide) • somewhat equivalent to toff • thus both need to be taken into account simultaneously • Question: what’s more important? • For the ice properties: • fSAM=0.0, toff=0 is mostly same as SPICE2x • fSAM=1.0, toff=0 fits well, see plots (in ps-SAM-only) at http://icecube.wisc.edu/~dima/work/IceCube-ftp/ppc/try/ • fSAM=0.0, toff allowed to vary: fits well, resulting toff~35 ns • b) For the use with muons: • delta-T plots: wait for results from Jake
Dependence on g=<cos(q)> and fSAM g=<cos(q)> fSAM 0.8 0 0.9 0 0.95 0 0.9 0.3 0.9 0.5 0.9 1.0 flashing 63-50 64-50 64-51 64-52 64-49 64-48 72-50
Systematics checks • Q: hole vs. nominal angular sensitivity model: • does the hole ice model fit well delta-T distributions well because it was used in fitting the ice? • A: no. • 1. See plots from the SPICE talk at Annapolis meeting (also next page), there is no difference between ice properties with ice fits run with hole ice vs. nominal ice. • 2. The SPICE2x used by Jake was also fitted with hole ice model, but, the agreement in delta-T was worse (and could not be fixed by changing the bubble density by as much as ~ 2 - 3 times).
Ratio to SPICE2x 7% uncertainty 5% uncertainty py=2.1 py=3.1