LSND versus MiniBooNE: Sterile neutrinos with energy dependent masses and mixing?

Standard active--sterile neutrino oscillations do not provide a satisfactory description of the LSND evidence for neutrino oscillations together with the constraints from MiniBooNE and other null-result short-baseline oscillation experiments. However, if the mass or the mixing of the sterile neutrino depends in an exotic way on its energy all data become consistent. I explore the phenomenological consequences of the assumption that either the mass or the mixing scales with the neutrino energy as $1/E_\nu^r (r > 0)$. Since the neutrino energy in LSND is about 40 MeV, whereas MiniBooNE operates at around 1 GeV, oscillations get suppressed in MiniBooNE and the two results become fully compatible for $r\gtrsim 0.2$. Furthermore, also the global fit of all relevant data improves significantly by exploring the different energy regimes of the various experiments. The best fit $\chi^2$ decreases by 12.7 (14.1) units with respect to standard sterile neutrino oscillations if the mass (mixing) scales with energy.