Testing neutrino spectra formation in collapsing stars with the diffuse supernova neutrino flux

I address the question of what can be learned from the observation of the diffuse supernova neutrino flux in the precision phase, at next generation detectors of Megaton scale. An analytical study of the spectrum of the diffuse flux shows that, above realistic detection thresholds of 10 MeV or higher, the spectrum essentially reflects the exponential-times-polynomial structure of the original neutrino spectrum at the emission point. There is only a weak (tens of percent) dependence on the power β describing the growth of the supernova rate with the redshift. Different original neutrino spectra correspond to large differences in the observed spectrum of events at a water Cherenkov detector: for typical supernova rates, the ratio of the numbers of events in the first and second energy bins (of 5 MeV width) varies in the interval 1.5-4.3 for pure water (energy threshold 18 MeV) and in the range 1-2.5 for water with gadolinium (10 MeV threshold). In the first case, discrimination would be difficult due to the large errors associated with background. With gadolinium, instead, the reduction of the total error down to the 10%-20% level would allow spectral sensitivity, with a dramatic improvement of precision with respect to the SN1987A data. Even in this latter case, for typical neutrino luminosity the dependence on β is below sensitivity, so that it can be safely neglected in data analysis.