The 2-3 mixing and mass split: atmospheric neutrinos and magnetized spectrometers

We study dependence of the atmospheric $\nu_\mu$ and $\bar{\nu}_\mu$ fluxes on the deviations of the 2-3 mixing from maximal, $|45^\circ - \theta_{23}|$, on the $\theta_{23}$-octant and on the neutrino mass splitting $\Delta m_{32}^2$. Analytic expressions for the $\theta_{23}-$deviation effect and the octant asymmetry are derived. We present conservative estimations of sensitivities of the iron (magnetized) calorimeter detectors (ICAL) to these parameters. ICAL can establish the $\theta_{23}$-deviation at higher than 3$\sigma$ confidence level if $| 45^\circ - \theta_{23}| > 6^{\circ}$ with the exposure of 1 Mton$\cdot$yr. Sensitivity to the octant is low for zero or very small 1-3 mixing, but it can be substantially enhanced for $\theta_{13} > 3^\circ$. ICAL can measure the difference of $\Delta m_{32}^2$ in $\nu$ and $\bar\nu$ channels (the CPT test) with accuracy $0.8\times 10^{-4}$ eV$^2$ (3$\sigma$) with 1 Mton$\cdot$yr exposure, and the present MINOS result can be excluded at $>5\sigma$ confidence level. We discuss possible ways to further improve sensitivity of the magnetized spectrometers.