All distributions show downgoing muons (minbias data stream, prescale of 1/200) at trigger level or with a paraboloid Sigma<2.5deg and Ndir>=5 cut. All data and sim were processed identically using the std-processing meta-project. The IC22 data is from Sept 28 and 29, 2007. The simulation comes from corsika 750, coincident 630, and nugen_numu 752. You can view the descriptions on the simulation webpage. All plots are scaled to the livetime. For each variable, I give the plot in linear and log, followed by the exp/sim bin-by-bin ratio in linear and log also. For a fair comparison, I had to require that enough DOMs are hit for the IceCube SMT so that TWR and IceTop events are not included.
The goal is to show comparisons for relevant cut variables at level2 both at trigger level and with some possible analysis-level cuts.
If the distributions agree for good down-going muons, we can be reasonably
confident that we understand how our cuts will perform on well-reconstructed
The generic analysis-level cuts I chose were:
Paraboloid Sigma < 2.5 deg
NdirC >= 5
These cuts won't quite get us down to a pure neutrino sample, but they do reduce the misreconstructed rate to ~ atms neutrino rate. The neutrinos have a median angular resolution of 1.7deg.
Bayesian Likelihood Ratio
It is worth mentioning that there is a known bug in the photonics tables in this simulation. The impact is still being evaluated.