Neutrino Superbeams and the Magic Baseline

We examine the sensitivity to ν µ → ν e of a conceptual experiment with a neutrino superbeam incident on a Megaton-scale water Cherenkov detector over a " magic " baseline ∼7300 km. With realistic beam intensity and exposure , the experiment may unambiguously probe sin 2 2θ 13 and the sign of ∆m 2 31 down to sin 2 2θ 13 ∼ 10 −3. Detecting the subdominant oscillation ν µ → ν e on the " atmospheric " scale of L/E has emerged as a priority for long-baseline accelerator experiments. This is because the ν µ → ν e and ¯ ν µ → ¯ ν e probabilities are sensitive to yet-unknown parameters of neutrino mixing: the mixing angle θ 13 , the sign of the " atmospheric " mass-squared difference ∆m 2 31 , and the CP-violating phase δ CP [1]. However, extracting the values of these parameters from measured probabilities will encounter the problem of degenerate solutions [2]. In particular, the asymmetry between P (ν µ → ν e) and P (¯ ν µ → ¯ ν e) may arise from either the intrinsic CP violation and the matter effect 1 that is correlated with the sign of ∆m 2 31 [3]. The degeneracies can be resolved by comparing the data taken with a shorter and longer baselines [4]. Selecting the latter as the " magic " baseline L magic ≃ 7300 km will render this strategy particularly effective: for L = L magic , all ∆m 2 21-induced effects like CP violation are predicted to vanish up to second order of the small parameter ∆m 2 21 /∆m 2 31 [2, 5]. Therefore, selecting L = L magic may allow to uniquely determine sin 2 2θ 13 and the sign of ∆m 2 31 , but not δ CP which should be probed with a shorter baseline. In this paper, we discuss a conceptual experiment that involves a neutrino " superbeam " incident on a water Cherenkov detector over a magic baseline of L = 7340 km 1. A water Cherenkov target is selected on the merit of good separation and spectrometry of electromagnetic showers [6], and is assumed to be a megaton-scale detector like UNO or Hyper-Kamiokande [7]. In tuning the energy of the neutrino beam, one must take into account that the E ν-dependence of oscillation probability for L = 7340 km is strongly …