2010ApJ...718..502M

In recent years, analyses of eclipsing binary systems have unveiled differences between the observed fundamental properties of low-mass stars and those predicted by stellar structure models. Particularly, radius and effective temperatures computed from models are ∼5%-10% lower and ∼3%-5% higher than observed, respectively. These discrepancies have been attributed to different factors, notably the high levels of magnetic activity present on these stars. In this paper, we test the effect of magnetic activity both on models and on the observational analysis of eclipsing binaries using a sample of such systems with accurate fundamental properties. Regarding stellar models, we have found that unrealistically high spot coverages need to be assumed to reproduce the observations. Tests considering metallicity effects and missing opacities on models indicate that these are not able to explain the radius discrepancies observed. With respect to the observations, we have tested the effect of several spot distributions on the light curve analysis. Our results show that spots cause systematic deviations on the stellar radii derived from light curve analysis when mainly distributed over the stellar poles. Assuming the existence of polar spots, overall agreement between models and observations is reached when ∼35% spot coverage is considered on stellar models. Such spot coverage induces a systematic deviation in the radius determination from the light curve analysis of ∼3% and is also compatible with the modulations observed on the light curves of these systems. Finally, we have found that the effect of activity or rotation on convective transport in partially radiative stars may also contribute to the explanation of the differences seen in some of the systems with shorter orbital periods.