AT2017gfo: Bayesian inference and model selection of multi-component kilonovae and constraints on the neutron star equation of state

TL;DR

This paper uses Bayesian inference on the kilonova AT2017gfo to determine ejecta properties, favor multi-component models, and constrain neutron star parameters, providing insights into the neutron star equation of state.

Contribution

It introduces a Bayesian framework for modeling kilonovae with multi-component, anisotropic ejecta, improving constraints on neutron star properties from observational data.

Findings

Anisotropic, multi-component models are strongly favored over spherical ones.

The best model includes dynamical ejecta, neutrino, and viscous winds components.

Constraints on binary mass ratio and neutron star radius are derived from the model.

Abstract

The joint detection of the gravitational wave GW170817, of the short $\gamma$-ray burst GRB170817A and of the kilonova AT2017gfo, generated by the the binary neutron star merger observed on August 17, 2017, is a milestone in multimessenger astronomy and provides new constraints on the neutron star equation of state. We perform Bayesian inference and model selection on AT2017gfo using semi-analytical, multi-components models that also account for non-spherical ejecta. Observational data favor anisotropic geometries to spherically symmetric profiles, with a log-Bayes' factor of ${\sim}10^{4}$, and favor multi-component models against single-component ones. The best fitting model is an anisotropic three-component composed of dynamical ejecta plus neutrino and viscous winds. Using the dynamical ejecta parameters inferred from the best-fitting model and numerical-relativity relations connecting the ejecta properties to the binary properties, we constrain the binary mass ratio to $q<1.54$ and the reduced tidal parameter to $120<\tilde\Lambda<1110$. Finally, we combine the predictions from AT2017gfo with those from GW170817, constraining the radius of a neutron star of $1.4~{\rm M}_\odot$ to $12.2\pm0.5~{\rm km}$ ($1\sigma$ level). This prediction could be further strengthened by improving kilonova models with numerical-relativity information.