Gaia in Technicolor. I. A Simple Variability-Encoded Color-Magnitude Diagram

Abstract

The Hertzsprung-Russell diagram is one of the most important tools in stellar astrophysics. In its modern, empirical form as the Gaia color-magnitude diagram (CMD), each star is represented with a point determined from its time-averaged brightness and color. However, many stars are intrinsically variable, and the fixed-point representation ignores the variety of ways in which a star can vary. As a system pulsates, eclipses, is irradiated, erupts, or otherwise varies, it moves through the CMD along a trajectory whose size, slope, and shape are determined by the physics driving the brightness and color variability. Here we exploit the near-simultaneous, three-band, multi-epoch photometry (MEP) released for ~11.7 million sources in Gaia DR3 to construct a variability-encoded CMD (VECMD). In this updated take on the traditional CMD, each star is represented not by a point but by a linear segment approximating the path it traces in absolute magnitude and color as it varies. Because these trajectories are built from the raw epoch data, they are agnostic to period and do not require phase-folding. We show that radial-mode pulsators trace looped, Lissajous-like trajectories that can encode the phase lag between radius and temperature. The Hertzsprung bump is even visible in the CMD loops of some Cepheids. Variable binary systems can trace tight color-correlated lines and/or hooked paths. Accreting and dust-obscured systems trace long, high-amplitude tracks aligned with the outburst or reddening vector. Distinct regions of the VECMD are dominated by distinct trajectory morphologies. We argue that these trajectories are period-independent fingerprints well suited to machine learning classification, which we will pursue on the much larger epoch-photometry sample expected in Gaia DR4. Finally, we present a free app called IVECMD for finding, classifying, and analyzing variable stars in the VECMD.