Modeling the Gaia Color-Magnitude Diagram with Bayesian Neural Flows to Constrain Distance Estimates
This work provides significantly improved distance estimates for stars, aiding astronomers in studying Milky Way structure and substructure detection, though it is incremental as it builds on existing normalizing flow methods applied to new astronomical data.
The paper tackled the problem of estimating distances to stars from noisy Gaia data by developing a Bayesian neural flow model that learns a flexible color-magnitude diagram, resulting in a catalog of 640 million photometric distance posteriors with an average 48% improvement in precision over raw Gaia data and enhanced accuracy in noisy regimes.
We demonstrate an algorithm for learning a flexible color-magnitude diagram from noisy parallax and photometry measurements using a normalizing flow, a deep neural network capable of learning an arbitrary multi-dimensional probability distribution. We present a catalog of 640M photometric distance posteriors to nearby stars derived from this data-driven model using Gaia DR2 photometry and parallaxes. Dust estimation and dereddening is done iteratively inside the model and without prior distance information, using the Bayestar map. The signal-to-noise (precision) of distance measurements improves on average by more than 48% over the raw Gaia data, and we also demonstrate how the accuracy of distances have improved over other models, especially in the noisy-parallax regime. Applications are discussed, including significantly improved Milky Way disk separation and substructure detection. We conclude with a discussion of future work, which exploits the normalizing flow architecture to allow us to exactly marginalize over missing photometry, enabling the inclusion of many surveys without losing coverage.