Building and Calibrating the Binary Star Population Using Kepler Data

TL;DR

This paper develops a forward-modeling approach to simulate and calibrate the binary star population in the Kepler field, aligning synthetic models with observed eclipsing binaries to better understand star formation and evolution.

Contribution

It introduces a new method for modeling binary populations by iteratively fitting synthetic data to Kepler observations, incorporating observational constraints for improved accuracy.

Findings

Binary period distribution is flat above ~3.2 days.

Synthetic binary populations can be tuned to match Kepler eclipsing binary data.

The model provides a calibrated underlying binary star population.

Abstract

Modeling binary star populations is critical to linking the theories of star formation and stellar evolution with observations. In order to test these theories, we need accurate models of observable binary populations. The Kepler Eclipsing Binary Catalog (KEBC), with its estimated $>$90% completeness, provides an observational anchor on binary population models. In this work we present the results of a new forward-model of the binary star population in the Kepler field. The forward-model takes a single star population from a model of the galaxy and pairs the stars into binaries by applying the constraints on the population from the results of observational binary population surveys such as arXiv:1007.0414 and arXiv:1303.3028. A synthetic binary population is constructed from the initial distributions of orbital parameters. We identify the eclipsing binary sample from the generated binary star population and compare this with the observed sample of eclipsing binaries contained in the KEBC. Finally, we update the distributions of the synthetic population and repeat the process until the synthetic eclipsing binary sample agrees with the KEBC. The end result of this process is a model of the underlying binary star population that has been fit to observations. We find that for fixed flat mass ratio and eccentricity input distributions, the binary period distribution is logarithmically flat above $\sim$3.2d. With additional constraints on distributions from observations, we can further adjust the synthetic binary population by relaxing other input constraints, such as mass ratio and eccentricity.