We present interferometric angular diameter measurements of 21 low-mass, Kand M-dwarfs made with the CHARA Array. This sample is enhanced by adding a collection of radii measurements published in the literature to form a total data set of 33 K-M dwarfs with diameters measured to better than 5%. We use these data in combination with the Hipparcos parallax and new measurements of the star's bolometric flux to compute absolute luminosities, linear radii, and effective temperatures for the stars. We develop empirical relations for ∼K0 to M4 mainsequence stars that link the stellar temperature, radius, and luminosity to thebroad-band color index and stellar metallicity [Fe/H]. These relations are valid for metallicities ranging from [Fe/H] = −0.5 to +0.1 dex, and are accurate to ∼2%, ∼5%, and ∼4% for temperature, radius, and luminosity, respectively. Our results show that it is necessary to use metallicity dependent transformations in order to properly convert colors into stellar temperatures, radii, and luminosities. Alternatively, we find no sensitivity to metallicity on relations we construct to the global properties of a star omitting color information e.g., temperature-radius and temperatureluminosity. Thus, we are able to empirically quantify to what order the star's observed color index is impacted by the stellar iron abundance. In addition to the empirical relations, we also provide a representative look-up table via stellar spectral classifications using this collection of data. Robust examinations of single star temperatures and radii compared to evolutionary model predictions on the luminosity -temperature and luminosity -radius planes reveals that models overestimate the temperatures of stars with surface temperatures < 5000 K by ∼ 3%, and underestimate the radii of stars with radii < 0.7 R ⊙ by ∼ 5%. These conclusions additionally suggest that the models over account for the effects that the stellar metallicity may have on the astrophysical properties of an object. By comparing the interferometrically measured radii for the single star population to those of eclipsing binaries, we find that for a given mass, single and binary star radii are indistinguishable. However, we also find that for a given radius, the literature temperatures for binary stars are systematically lower compared to our interferometrically derived temperatures of single stars by ∼ 200 to 300 K. The nature of this offset is dependent on the validation of binary star temperatures; where bringing all measurements to a uniform and correctly calibrated temperature scale is needed to identify any influence stellar activity may have on the physical properties of a star. Lastly, we present a empirically determined HR diagram using fundamental properties presented here in combination with those in Boyajian et al. (2012) for a total of 74 nearby, main-sequence, A-to M-type stars, and define regions of habitability for the potential existence of sub-stellar mass companions in each system.

Astrometric measurements for 25 red dwarf systems are presented, including the first definitive trigonometric parallaxes for 20 systems within 10 pc of the Sun, the horizon of the RECONS sample. The three nearest systems that had no previous trigonometric parallaxes (other than perhaps rough preliminary efforts) are SO 0253+1652 (3:84 AE 0:04 pc, the 23rd nearest system), SCR 1845À6357 AB (3:85 AE 0:02 pc, 24th nearest), and LHS 1723 (5:32 AE 0:04 pc, 56th nearest). In total, seven of the systems reported here rank among the nearest 100 stellar systems. Supporting photometric and spectroscopic observations have been made to provide full characterization of the systems, including complete VRIJHK s photometry and spectral types. A study of the variability of 27 targets reveals six obvious variable stars, including GJ 1207, for which we observed a flare event in the V band that caused it to brighten by 1.7 mag. Improved parallaxes for GJ 54 AB and GJ 1061, both important members of the 10 pc sample, are also reported. Definitive parallaxes for GJ 1001 A, GJ 633, and GJ 2130 ABC, all of which have been reported to be within 10 pc, indicate that they are beyond 10 pc. From the analysis of systems with (previously) high trigonometric parallax errors, we conclude that parallaxes with errors in excess of 10 mas are insufficiently reliable for inclusion in the RECONS sample. The cumulative total of new additions to the 10 pc sample since 2000 is now 34 systems: 28 by the RECONS team and six by other groups. This total represents a net increase of 16% in the number of stellar systems reliably known to be nearer than 10 pc.