2018ApJ...856...10J

Theoretical stellar evolution models are constructed and tailored to the best known, observationally derived characteristics of metal-poor ([Fe/H] ∼ -2.3) stars representing a range of evolutionary phases: subgiant HD 140283, globular cluster M92, and four single, main sequence stars with well-determined parallaxes: HIP 46120, HIP 54639, HIP 106924, and WOLF 1137. It is found that the use of a solar-calibrated value of the mixing length parameter α_MLT in models of these objects is ineffective at reproducing their observed properties. Empirically calibrated values of α_MLT are presented for each object, accounting for uncertainties in the input physics employed in the models. It is advocated that the implementation of an adaptive mixing length is necessary in order for stellar evolution models to maintain fidelity in the era of high-precision observations.