Spallation Backgrounds in Super-Kamiokande Are Made in Muon-Induced Showers

Crucial questions about solar and supernova neutrinos remain unanswered. Super-Kamiokande has the exposure needed for progress, but detector backgrounds are a limiting factor. A leading component is the beta decays of isotopes produced by cosmic-ray muons and their secondaries, which initiate nuclear spallation reactions. Cuts of events after and surrounding muon tracks reduce this spallation decay background by $simeq 90%$ (at a cost of $simeq 20%$ deadtime), but its rate at 6--18 MeV is still dominant. A better way to cut this background was suggested in a Super-Kamiokande paper [Bays {it et al.}, Phys.~Rev.~D {bf 85}, 052007 (2012)] on a search for the diffuse supernova neutrino background. They found that spallation decays above 16 MeV were preceded near the same location by a peak in the apparent Cherenkov light profile from the muon; a more aggressive cut was applied to a limited section of the muon track, leading to decreased background without increased deadtime. We put their empirical discovery on a firm theoretical foundation. We show that almost all spallation decay isotopes are produced by muon-induced showers and that these showers are rare enough and energetic enough to be identifiable. This is the first such demonstration for any detector. We detail how the physics of showers explains the peak in the muon Cherenkov light profile and other Super-K observations. Our results provide a physical basis for practical improvements in background rejection that will benefit multiple studies. For solar neutrinos, in particular, it should be possible to dramatically reduce backgrounds at energies as low as 6 MeV.

First calculation of cosmic-ray muon spallation backgrounds for MeV astrophysical neutrino signals in Super-Kamiokande

When muons travel through matter, their energy losses lead to nuclear breakup (spallation) processes. The delayed decays of unstable daughter nuclei produced by cosmic-ray muons are important backgrounds for low-energy astrophysical neutrino experiments, e.g., those seeking to detect solar neutrino or Diffuse Supernova Neutrino Background (DSNB) signals. Even though Super-Kamiokande has strong general cuts to reduce these spallation-induced backgrounds, the remaining rate before additional cuts for specific signals is much larger than the signal rates for kinetic energies of about 6 -- 18 MeV. Surprisingly, there is no published calculation of the production and properties of these backgrounds in water, though there are such studies for scintillator. Using the simulation code FLUKA and theoretical insights, we detail how muons lose energy in water, produce secondary particles, how and where these secondaries produce isotopes, and the properties of the backgrounds from their decays. We reproduce Super-Kamiokande measurements of the total background to within a factor of 2, which is good given that the isotope yields vary by orders of magnitude and that some details of the experiment are unknown to us at this level. Our results break aggregate data into component isotopes, reveal their separate production mechanisms, and preserve correlations between them. We outline how to implement more effective background rejection techniques using this information. Reducing backgrounds in solar and DSNB studies by even a factor of a few could help lead to important new discoveries.