Mass hierarchy determination using neutrinos from multiple reactors

We report the results of Monte Carlo simulations of a medium baseline reactor neutrino experiment. The difference in baselines resulting from the 1 km separations of Daya Bay and Ling Ao reactors reduces the amplitudes of 1-3 oscillations at low energies, decreasing the sensitivity to the neutrino mass hierarchy. A perpendicular detector location eliminates this effect. We simulate experiments under several mountains perpendicular to the Daya Bay/Ling Ao reactors, considering in particular the background from the TaiShan and YangJiang reactor complexes. In general the hierarchy can be determined most reliably underneath the 1000 meter mountain BaiYunZhang, which is 44.5 km from Daya Bay. If some planned reactors are not built then near by 700 meter mountains at 47-51 km baselines gain a small advantage. Neglecting their low overhead burdens, hills near BaiMianShi or DongKeng would be the optimal locations. We use a weighted Fourier transform to avoid a spurious dependence on the high energy neutrino spectrum and find that a neural network can extract quantities which determine the hierarchy marginally better than the traditional RL + PV.