2007ApJ...660..732L

The recent increase in the number of radius measurements of very low mass stars from eclipsing binaries and interferometry of single stars has raised more questions about what could be causing the discrepancy between the observed radii and those predicted by models. The two main explanations being proposed are a correlation between the radii of the stars and either their activity levels or their metallicities. This paper presents a study of such correlations using all the data published to date. The study also investigates correlations between the radius deviations from the models and the masses of the stars. There is no clear correlation between activity level and radius for the single stars in the sample. These single stars are slow rotators, with typical velocities v_rotsini<3.0 km/s. A clear correlation however exists in the case of the faster rotating members of binaries. This result is based on the X-ray emission levels of the stars. There also appears to be an increase in the deviation of the radii of single stars from the models as a function of metallicity, as previously indicated by Berger et al. The stars in binaries do not seem to follow the same trend. Finally, the Baraffe et al. models reproduce well the radius observations below 0.30-0.35 M_☉, where the stars become fully convective, although this result is preliminary since almost all the sample stars in that mass range are slow rotators and metallicities have not been measured for most of them. The results indicate that stellar activity and metallicity play an important role in determining the radius of very low mass stars, at least above 0.35 M_☉.