2006ApJ...653..647D

We describe an automated method for assigning the most probable physical parameters to the components of an eclipsing binary, using only its photometric light curve and combined colors. With traditional methods, one attempts to optimize a multiparameter model over many iterations, so as to minimize the χ² value. We suggest an alternative method, in which one selects pairs of coeval stars from a set of theoretical stellar models and compares their simulated light curves and combined colors with the observations. This approach greatly reduces the parameter space over which one needs to search and allows one to estimate the components' masses, radii, and absolute magnitudes, without spectroscopic data. We have implemented this method in an automated program using published theoretical isochrones and limb-darkening coefficients. Since it is easy to automate, this method lends itself to systematic analyses of data sets consisting of photometric time series of large numbers of stars, such as those produced by OGLE, MACHO, TrES, HAT, and many others surveys.