Paper detail

Can a ``natural'' three-generation neutrino mixing scheme satisfy everything?

We examine the potential for a ``natural'' three-neutrino mixing scheme to satisfy available data and astrophysical arguments. By ``natural'' we mean no sterile neutrinos, and a neutrino mass hierarchy similar to that of the charged leptons. We seek to satisfy (or solve): 1. Accelerator and reactor neutrino oscillation constraints, including LSND; 2. The atmospheric muon neutrino deficit problem; 3. The solar neutrino problem; 4. Supernova $r$-process nucleosynthesis in neutrino-heated supernova ejecta; 5. Cold+hot dark matter models. We argue that putative supernova $r$-process nucleosynthesis bounds on two-neutrino flavor mixing can be applied directly to three-neutrino mixing in the case where one vacuum neutrino mass eigenvalue difference dominates the others. We show that in this ``one mass scale dominance'' limit, a natural three-neutrino oscillation solution meeting all the above constraints exists only if the atmospheric neutrino data {\em and} the LSND data can be explained with one neutrino mass difference. In this model, an explanation for the solar neutrino data can be effected by employing the {\em other} independent neutrino mass difference. Such a solution is only marginally allowed by the current data, and proposed long-baseline neutrino oscillation experiments can definitively rule it out. If it were ruled out, the simultaneous solution of the above constraints by neutrino oscillations would then require sterile neutrinos and/or a neutrino mass hierarchy of a different nature than that of the charged leptons.