[Paper Review] The Gaia-Kepler Stellar Properties Catalog. I. Homogeneous Fundamental Properties for 186,301 Kepler Stars

Travis A. Berger, Daniel Huber+4 more|arXiv (Cornell University)|Jan 21, 2020

Stellar, planetary, and galactic studies4 citations

TL;DR

This paper presents the Gaia-Kepler Stellar Properties Catalog, a homogeneous analysis of 186,301 Kepler stars using Gaia DR2 parallaxes, broadband photometry (g-Ks), and spectroscopic metallicities. By applying isochrone fitting calibrated on interferometric angular diameters, it achieves median uncertainties of 112 K in effective temperature, 0.05 dex in surface gravity, and 7% in mass, enabling precise stellar and exoplanet population studies across mass and age.

ABSTRACT

An accurate and precise Kepler Stellar Properties Catalog is essential for the interpretation of the Kepler exoplanet survey results. Previous Kepler Stellar Properties Catalogs have focused on reporting the best-available parameters for each star, but this has required combining data from a variety of heterogeneous sources. We present the Gaia-Kepler Stellar Properties Catalog, a set of stellar properties of 186,301 Kepler stars, homogeneously derived from isochrones and broadband photometry, Gaia Data Release 2 parallaxes, and spectroscopic metallicities, where available. Our photometric effective temperatures, derived from $g-K_s$ colors, are calibrated on stars with interferometric angular diameters. Median catalog uncertainties are 112 K for $T_{\\mathrm{eff}}$, 0.05 dex for $\\log g$, 4% for $R_\\star$, 7% for $M_\\star$, 13% for $\ ho_\\star$, 10% for $L_\\star$, and 56% for stellar age. These precise constraints on stellar properties for this sample of stars will allow unprecedented investigations into trends in stellar and exoplanet properties as a function of stellar mass and age. In addition, our homogeneous parameter determinations will permit more accurate calculations of planet occurrence and trends with stellar properties.

Motivation & Objective

• To create a homogeneous, precise, and accurate stellar parameter catalog for all Kepler stars to improve exoplanet occurrence rate calculations and stellar population studies.
• To overcome limitations of prior catalogs that combined heterogeneous data sources with inconsistent uncertainties.
• To leverage Gaia DR2 parallaxes, g-Ks photometry, and spectroscopic metallicities to derive fundamental stellar properties via isochrone fitting.
• To calibrate photometric effective temperatures using interferometric angular diameters for improved accuracy.
• To enable systematic investigations of stellar and exoplanet properties as functions of mass and age using a consistent, high-precision dataset.

Proposed method

• Used Gaia DR2 parallaxes for distance and luminosity constraints on 186,301 Kepler stars.
• Applied g-Ks broadband photometry to derive effective temperatures, calibrated on stars with interferometric angular diameters.
• Incorporated spectroscopic metallicities where available to improve isochrone fitting.
• Performed isochrone placement using evolutionary models to derive log g, mass, radius, density, luminosity, and age.
• Calculated uncertainties via posterior sampling and goodness-of-fit metrics, with 4σ model selection.
• Used machine-readable output with likelihood-based confidence intervals and terminal age of main sequence (TAMS) indicators.

Experimental results

Research questions

• RQ1 What are the most precise and homogeneous fundamental stellar properties for the full Kepler sample of 186,301 stars?
• RQ2 How do uncertainties in stellar parameters (T_eff, log g, R_*, M_*, ρ_*, L_*, age) compare across different methods and data combinations?
• RQ3 To what extent does incorporating Gaia DR2 parallaxes and isochrone fitting improve the accuracy of stellar mass and radius estimates compared to photometric-only methods?
• RQ4 How do the derived stellar properties correlate with exoplanet occurrence rates and host star characteristics across mass and age?
• RQ5 What fraction of Kepler stars are main sequence, subgiant, or giant, and how does this distribution compare to previous estimates?

Key findings

• Median uncertainty in effective temperature (T_eff) is 112 K, calibrated using interferometric angular diameters.
• Median uncertainty in surface gravity (log g) is 0.05 dex, significantly improving on prior photometric-only estimates.
• Median uncertainty in stellar mass (M_*) is 7%, and in radius (R_*) is 4%, reflecting high precision from isochrone fitting with Gaia parallaxes.
• Median uncertainty in stellar density (ρ_*) is 13%, and in luminosity (L_*) is 10%, with age uncertainty at 56% due to degeneracies in isochrone placement.
• The catalog identifies 67% of Kepler stars as main sequence, 21% as subgiants, and 12% as giants, consistent with a homogeneous analysis.
• Ages are flagged as unreliable if goodness-of-fit (GOF) < 0.99 or TAMS > 20 Gyr, ensuring robustness in age estimates.