Spectroscopic and eclipsing binary systems offer the best means for determining accurate physical properties of stars, including their masses and radii. The data available for low-mass stars have yielded firm evidence that stellar structure models pr
edict smaller radii and higher effective temperatures than observed, but the number of systems with detailed analyses is still small. In this paper we present a complete reanalysis of one of such eclipsing systems, CM Dra, composed of two dM4.5 stars. New and existing light curves as well as a radial velocity curve are modeled to measure the physical properties of both components. The masses and radii determined for the components of CM Dra are M1=0.2310+/-0.0009 Msun, M2=0.2141+/-0.0010 Msun, R1=0.2534+/-0.0019 Rsun, and R2=0.2396+/-0.0015 Rsun. With relative uncertainties well below the 1% level, these values constitute the most accurate properties to date for fully convective stars. This makes CM Dra a valuable benchmark for testing theoretical models. In comparing our measurements with theory, we confirm the discrepancies reported previously for other low-mass eclipsing binaries. These discrepancies seem likely to be due to the effects of magnetic activity. We find that the orbit of this system is slightly eccentric, and we have made use of eclipse timings spanning three decades to infer the apsidal motion and other related properties.
Recent analyses of low-mass eclipsing binary stars have unveiled a significant disagreement between the observations and the predictions of stellar structure models. Results show that theoretical models underestimate the radii and overestimate the ef
fective temperatures of low-mass stars but yield luminosities that accord with observations. A hypothesis based upon the effects of stellar activity was put forward to explain the discrepancies. In this paper we study the existence of the same trend in single active stars and provide a consistent scenario to explain systematic differences between active and inactive stars in the H-R diagram reported earlier. The analysis is done using single field stars of spectral types late-K and M and computing their bolometric magnitudes and temperatures through infrared colours and spectral indices. The properties of the stars in samples of active and inactive stars are compared statistically to reveal systematic differences. After accounting for a number of possible bias effects, active stars are shown to be cooler than inactive stars of similar luminosity therefore implying a larger radius as well, in proportions that are in excellent agreement with those found from eclipsing binaries. The present results generalise the existence of strong radius and temperature dependences on stellar activity to the entire population of low-mass stars, regardless of their membership in close binary systems.
