Title: Retrieval of energy spectra for all flavor of neutrinos from core-collapse supernova with multiple detectors

Abstract: We present a new method by which to retrieve energy spectrum for all flavor of neutrinos from core-collapse supernova (CCSN). In the retrieval process, we do not assume any analytic formulae to express the energy spectrum of neutrinos but rather take a direct way of spectrum reconstruction from the observed data; the Singular Value Decomposition algorithm with a newly developed adaptive energy-gridding technique is adopted. We employ three independent reaction channels having different flavor sensitivity to neutrinos. Two reaction channels, inverse beta decay on proton and elastic scattering on electrons, from a water Cherenkov detector such as Super-Kamiokande (SK) and Hyper-Kamiokande (HK), and a charged current reaction channel with Argon from the Deep Underground Neutrino Experiment (DUNE) are adopted. Given neutrino oscillation models, we iteratively search the neutrino energy spectra at the CCSN source until they provide the consistent event counts in the three reaction channels. We test the capability of our method by demonstrating the spectrum retrieval to a theoretical neutrino data computed by our recent three-dimensional CCSN simulation. Although the energy spectrum with either electron-type or electron-type anti-neutrinos at the CCSN source has relatively large error compared to that of other species, the joint analysis with HK + DUNE or SK + DUNE will provide precise energy spectrum of all flavors of neutrinos at the source. Finally, we discuss perspectives for improvements of our method by using neutrino data of other detectors.