Markov Chain Monte Carlo Optimization applied to Dyson's Visual Double Stars

TL;DR

This paper applies Bayesian optimization with Markov Chain Monte Carlo methods to estimate orbital parameters of visual double stars, providing improved accuracy and uncertainty estimates, especially for long-period systems, and deriving dynamical masses using parallax data.

Contribution

It introduces a Bayesian MCMC approach for orbital parameter estimation of binary stars, offering unbiased uncertainties and improved long-period system analysis.

Findings

Agreement with Dyson for short-period systems

Enhanced estimates for long-period systems with additional data

Dynamical masses derived from parallax measurements

Abstract

Estimates of orbital parameters were made using a Bayesian optimization technique on astrometric data for 25 visual binary systems catalogued a century ago by the ninth Astronomer Royal, Sir Frank Dyson. An advantage of this method is that it provides reliable, unbiased uncertainty estimates for the optimized parameters. Reasonable agreement is found for the short period (< 100 yr) systems between the current study and Dyson, with superior estimation for the longer systems through the inclusion of an additional century of data. Dynamical masses are presented for the systems through the inclusion of parallax measurements.