Summary of IceCube Tau Neutrino Searches and Flavor Composition Measurements of the Diffuse Astrophysical Neutrino Flux

Abstract: We present a summary of the flavor composition measurements for the diffuse astrophysical neutrino flux using data from the IceCube Neutrino Observatory at the South Pole. IceCube has identified candidate astrophysical tau neutrinos through two different approaches. One approach used a dedicated particle identification algorithm for the classification and reconstruction of the 'Double Cascade' event topology, a signature of tau neutrino charged current interactions. This first approach is applied to the High Energy Starting Events (HESE) sample, an all-sky, all-flavor set of neutrino events with energy above 60~TeV encompassing 12 years of IceCube livetime. We show that the addition of more years of data and updated ice properties on the HESE sample delivers tighter constraints on the flavor composition of the astrophysical neutrino flux than previous IceCube analyses, in particular when it is fit in combination with high statistics samples of through-going tracks and cascades. A second approach uses a sensitive machine-learning-based selection technique that finds seven candidate events in 9.7 years of IceCube data. This approach excludes the zero astrophysical tau neutrino hypothesis at the highest statistical significance to date.