Detached eclipsing binaries (DEBs), especially double-lined systems, are uniquely positioned as high-quality benchmark stars due to the fact we can measure precise (<0.2%), model-independent masses and radii from light curves and radial velocities. However, effective temperature remains a limiting factor in their use as benchmarks: heterogeneity between methods and large systematic errors often plague the literature, despite small formal errors on individual measurements. I will present a new method to determine fundamental effective temperatures for DEBs using multi-band photometry, radii and Gaia parallaxes, and highlight results from a growing sample of well-characterised FGK-type systems.
Benchmark samples like these have a wide variety of applications, such as in the calibration of stellar models and development of data-driven pipelines. In the second part of the talk, I will introduce a neural-network-powered pipeline for deriving homogeneous stellar parameters and chemical abundances for M-dwarfs, as part of the upcoming PLATO mission. The pipeline combines photometry and H-band spectroscopy within a Bayesian framework, and requires extensive testing and validation against benchmark stars.
By combining these two approaches of (i) measuring the properties of a handful of benchmark stars in detail, and (ii) characterising large populations of stars with validated methods, we hope to broaden our understanding of the lives and processes of cool stars and their planets.