Improved analysis of SN1987A antineutrino events

TL;DR

This paper introduces a new parameterization of supernova antineutrino flux, analyzes SN1987A data comprehensively, and finds results consistent with standard core-collapse models, improving understanding of supernova neutrino emission.

Contribution

It presents a novel flux parameterization that aligns with supernova explosion models and applies a comprehensive analysis of SN1987A data considering all event characteristics without external priors.

Findings

Initial luminous phase (accretion) can be rejected at 2% significance.

Best-fit parameters agree with standard supernova models.

Total neutrino energy emitted is approximately 2.2×10^{53} erg.

Abstract

We propose a new parameterization of the antineutrino flux from core collapse supernovae, that allows an interpretation of its astrophysical parameters within the Bethe and Wilson scenario for the explosion, and that leads to a reasonable (smooth) behavior of the average energy and of the luminosity curve. We apply it to analyze the events observed by Kamiokande-II, IMB and Baksan detectors in correlation with SN1987A. For the first time, we consider in the same analysis all data characteristics: times, energies and angles of the observed events. We account for the presence of background and evaluate the impact of neutrino oscillations. The hypothesis that the initial luminous phase of emission (accretion) is absent can be rejected at the 2 % significance level. Without the need to impose external priors in the likelihood analysis, the best-fit values of the astrophysical parameters are found to be in remarkable agreement with the expectations of the standard core-collapse scenario; in particular, the electron antineutrino-sphere radius is 16 km, the duration of the accretion phase is found to be 0.55 s, and the initial accreting mass is 0.22 M(sun). Similarly the total energy emitted in neutrinos is 2.2*10^{53} erg, again close to the expectations. The errors on the parameters are evaluated and found to be relatively large, consistently with the limited number of detected events; the two dimensional confidence regions, that demonstrate the main correlations between the parameters, are also given.