Data

The run selection and hit cleaning were applied according to the standard scripts decided by the collaboration to be used in the LBNL filtering. They are shown in detail in appendix A. There is one exception. Run numbers 773 and 774 from day 190 and 786, 787, 788, 790 from day 195 have been removed because they contain laser data. The live-time for our data set relied on information extracted from a table shown in the internal report from Zeuthen [5] (Tables B13 to B17 in Appendix B). The live-time estimated is 138.2 days. The details are in appendix B.

After reducing the experimental data by 90% with the UCI filter, an energy reconstruction was applied using the minimization function called a_pt_phpnh in RECOOS. This function estimates the energy of an event based on the probabilities for photon detection. The energy reconstruction is a useful component in our analysis because the likelihood parameter of the energy fit is sensitive to the accuracy of the reconstructed trajectory. Note that this analysis does not directly utilize the reconstructed energy so there is no direct bias on the energy of the event. However, it is observed higher energy events are correlated with lower likelihood values (Figure 1).

  

Figure 1: likelihood parameter of the energy fit versus the muon energy (specifically the muon energy at the center of the detector). Shows higher energy muons tend to have better energy fits (smaller likelihood).