Improved neutrino point source search method for IceCube's high energy starting event sample

Abstract

High Energy Neutrino Astronomy is a new field that has the potential to solve the mystery of the origin of cosmic ray particles, the highest energy particles that have been observed. The IceCube Neutrino Observatory has recently discovered the existence of a high energy astrophysical neutrino flux. We are able to calculate for our "track-like" sub-population of events a directional origin to within a square degree, yet it remains difficult to establish whether these neutrinos come from bright distinguishable sources or a sea of individual weak neutrino producers. The current goal of neutrino point source analyses is the observation of a bright neutrino source above expected background rates of neutrinos, in our case both from a uniform distribution of astrophysical neutrino sources and background neutrino events produced in cosmic ray interactions with our own atmosphere. This thesis presents a new form of point source analysis that tests the data under the model of the combination of multiple background hypotheses and a single point source hypothesis, where the predicted background distributions can be produced through simulation. We apply this analysis to the High Energy Starting Event sample produced by IceCube, and show how this analysis differs to previous analyses on these events. We find that the fit using a background in the likelihood that does not match the distribution of the events can result in a bias in the fitted strength of a point source, and that the relative power of the analysis compared to the previous point source analysis can depend on the point source location, where the power is seen to be comparable in the southern sky but improved in the northern sky. Our results from applying our new form of point source analysis to the data do not show strong evidence for a point source hypothesis, with p-values of 0.468 for all events in our sample and 0.922 for a subset of shower topology events in our sample.