One of the important goals for future neutrino telescopes is to identify the flavors of astrophysical neutrinos and therefore determine the flavor ratio. The flavor ratio of astrophysical neutrinos observed on the Earth depends on both the initial fl
avor ratio at the source and flavor transitions taking place during propagations of these neutrinos. We propose a model independent parametrization for describing the above flavor transitions. A few flavor transition models are employed to test our parametrization. The observational test for flavor transition mechanisms through our parametrization is discussed.
We discuss the reconstruction of neutrino flavor ratios at astrophysical sources through the future neutrino-telescope measurements. Taking the ranges of neutrino mixing parameters $theta_{ij}$ as those given by the current global fit, we demonstrate
by a statistical method that the accuracies in the measurements of energy-independent ratios $Requivphi ( u_{mu})/(phi ( u_{e})+phi ( u_{tau}))$ and $Sequivphi ( u_e)/phi ( u_{tau})$ among integrated neutrino flux should both be better than 10% in order to distinguish between the pion source and the muon-damped source at the $3 sigma$ level. The 10% accuracy needed for measuring $R$ and $S$ requires an improved understanding on the background atmospheric neutrino flux to a better than 10% level in the future. We discuss the applicability of our analysis to practical situations that the diffuse astrophysical neutrino flux arises from different types of sources and each point source has a neutrino flavor ratio varying with energies. We also discuss the effect of leptonic CP phase on the flavor-ratio reconstruction.
