Lightcurves reproduced using current ASTERIA and modern SNEWPY yield greater numbers of hits

[sgriswol]

This figure was included with arXiv:2308.01843 and describes IceCube's signal lightcurve from a model by Nakazato.

An attempt to reproduce this plot was made using the apr25_models_xs_fixes branch of ASTERIA, Python 3.11, and SNEWPY 1.6b1. The plotting notebook is committed to that branch under docs/nb/model_phase_plots.ipynb.

The produced lightcurves report more hits than the original, though only when neutrino oscillations are enabled. Approximately 25% more signal hits are reported under the inverted hierarchy, and 10% under the normal hierarchy. Note that different background hit series are used (the seed used in the original figure is unknown, and was not fixed).

I can think of two potential causes for this difference, with the first being the most likely to account for the hierarchy-dependent variation.

1. Updates to flavor transformations and neutrino mixing parameters in SNEWPY. The original used ASTERIA flavor transformations (see python/asteria/oscillation.py) which only depended on theta12 (theta12 = 33.2 deg was chosen for the original figure).

2. Updates to deadtime efficiency. Commits 4fb0ff6 and d35772b updated the method by which deadtime efficiency was calculated.

Some investigation into this difference may be warranted.