The observation of neutrino flavor oscillations by independent measurements of the Sudbury Neutrino Observatory (SNO) and the Super Kamiokande projects demonstrated the existence of neutrino masses, providing the first evidence of physics beyond the standard model (SM).To complete the picture of the neutrino oscillation formalism, one of the most important open questions is the ordering of the three neutrino mass eigenstates ?1, ?2 and ?3. After incredible efforts by solar, atmospheric and accelerator neutrino experiments, this problem is now reduced to a simpler question: whether the third mass state ?3 is above or below the other two. The former case corresponds to the so called normal ordering (NO) whereas the latter is called inverted ordering (IO). So far, there is not enough sensitivity to the neutrino mass ordering (NMO) to reach a definitive conclusion on this key oscillation parameter. Sensitivity to the NMO in atmospheric neutrino oscillation experiments arises from distinct matter effects in the oscillation probabilities for neutrinos in NO or antineutrinos in IO, after propagation through the matter profile of the Earth. The other two combinations: antineutrinos in NO and neutrinos in IO, present negligible matter effects in their oscillation profiles. The IceCube neutrino observatory has performed a measurement on this parameter using the DeepCore detector, with results that are consistent with both orderings. A main challenge for NMO measurements with atmospheric neutrinos in IceCube is that sensitivity to the NMO is maximal when comparing detector expectations for NO/IO to experimental data for clean samples of neutrinos or antineutrinos, but this separation is typically not possible for very large-scale Cherenkov detectors. However, sensitivity to the NMO is achieved through statistical differences in the total neutrino + antineutrino count rates in the detector between orderings, arising from differences in the cross sections and atmospheric fluxes for neutrinos and antineutrinos. This work presents a study of the neutrino mass ordering with DeepCore that aims to directly separate ?? and ̄ ??, thus incorporating the oscillation matter effects to full potential sensitivity. The novel algorithm developed leverages the difference in signatures of muon decay for ?? and ̄ ?? within the natural ice sheet. Within the sparsely instrumented DeepCore detector array, the ability to directly detect these low-energy signatures and identify the decay lifetimes was implemented. Dominant backgrounds in the DeepCore optical modules, including artifacts in the operation of the photomultiplier tubes (dark noise and afterpulsing) and the onboard electronics, were studied in detail and improved techniques, including improved models, to mitigate the effects were incorporated in the study. A boosted decision tree was trained to optimize the ??/ ̄ ?? separation power with the other key input parameter (neutrino energy, direction) of the mass ordering analysis, resulting in a modest improvement in the signature sensitivity (0.358+0.060−0.057? compared to 0.226+0.072−0.068?). A binary ensemble test utilizing 7.5 years of DeepCore data demonstrated a small preference for the “normal” neutrino mass ordering over the inverted case with p-values ?NO = 0.72 and ?IO = 0.17. The ??? value of the result, including the ??/ ̄ ?? separation parameter, is 0.61, consistent with both mass ordering scenarios. In the future, the fully developed analysis will be enhanced with data from the IceCube Upgrade which will incorporate an array design that will make it possible to enhance the detectability of the muon decay signature photons while significantly reducing the impact of internal optical module backgrounds.
Read
