We investigate the sensitivity of some of the proposed next-generation neutrino experiments to a galactic supernova. In particular, we study how well the supernova parameters (the average energies and luminosities) can be separated from the unknown neutrino oscillation parameters ($��_{13}$ and the neutrino mass hierarchy). Three types of experiments, all in the 100-kton class, are compared. These are: a 540 kton water-Cherenkov detector, a 100 kton liquid Argon detector and a 50 kton scintillator detector. We demonstrate that practically all of these proposed detectors have the possibility to determine the hierarchy of the neutrino masses if the angle $��_{13}$ is sufficiently large ($\sin^2(��_{13}) > 10^{-4}$) and the hierarchy of the average energies is larger than about 20%. They can at the same time determine some of the supernova parameters well. The average energy of the $��_��$ and $��_��$ species can be determined within 5% uncertainty in most of the parameter space suggested by supernova simulations. The detection of several separable channels measuring different combinations of charged current and neutral current processes is crucial for determining the value of $��_{13}$ and the hierarchy. However, there are cases where a few of the SN parameters can be determined rather well even if only the main charged current detection channel is available.

44 pages, 13 figures. Appendix added and improvements in text. Matches published version