LP349-25AB: Application of Magneto-convection Model to an M Dwarf Binary Where Radio and Astrometric Data have Yielded Masses with 0.7% Precision

Abstract Reliable (±(1–2)%) estimates of stellar masses have typically been obtained from eclipsing binaries. In such cases, derived values of radii and T eff are in many cases discrepant relative to standard models, especially in active stars. These discrepancies led to the development of magneto-convection theory. Recently, an independent technique has emerged for determining precise masses of radio-emitting binary components: interferometric observations at multiple points around the orbit. S. Curiel et al. have used this technique to determine the masses of both components of an M8+M9 binary (LP349-25AB) with remarkably high precision: ±0.7%. However, applying standard models suggests that the components are not co-eval, differing in age by ∼30%. But LP349-25 is such a close binary (with a separation of 1.2 au) that non-coevality seems unrealistic. Here, we report on magneto-convective modeling which resolves the apparent non-coevality while simultaneously explaining why the primary component is variable, but the secondary component is not.