Discovering neutron stars with LISA via measurements of orbital eccentricity in Galactic binaries

TL;DR

This paper explores how LISA can identify neutron star binaries through orbital eccentricity measurements, using Bayesian analysis of simulated signals to improve source classification and understand binary evolution.

Contribution

It introduces a multi-harmonic heterodyning Bayesian approach to detect and analyze eccentric LISA signals, enabling neutron star identification and source reclassification.

Findings

LISA can detect eccentricity down to very small values.

Eccentricity helps distinguish neutron star binaries from white dwarf binaries.

Combining circular analyses can reclassify sources as eccentric binaries.

Abstract

LISA will detect $\sim \! 10^4$ Galactic binaries, the majority being double white dwarfs. However, approximately $\sim \! 1 \textrm{--} 5 \%$ of these systems will contain neutron stars which, if they can be correctly identified, will provide new opportunities for studying binary evolution pathways involving mass reversal and supernovae as well as being promising targets for multi-messenger observations. Eccentricity, expected from neutron star natal kicks, will be a key identifying signature for binaries containing a neutron star. Eccentric binaries radiate at widely-spaced frequency harmonics that must first be identified as originating from a single source and then analysed coherently. A multi-harmonic heterodyning approach for this type of data analysis is used to perform Bayesian parameter estimation on a range of simulated eccentric LISA signals. This is used to: (i) investigate LISA's ability to measure orbital eccentricity and to quantify the minimum detectable eccentricity; (ii) demonstrate how eccentricity and periastron precession help to break the mass degeneracy allowing the individual component masses to be inferred, potentially confirming the presence of a neutron star; (iii) investigate the possibility of source misidentification when the individual harmonics of an eccentric binary masquerade as separate circular binaries; and (iv) investigate the possibility of source reclassification, where parameter estimation results of multiple circular analyses are combined in postprocessing to quickly infer the parameters of an eccentric source. The broader implications of this for the ongoing design of the LISA global fit are also discussed.