2003ApJ...589.1020D

We present a new method for differentiating between planetary transits and eclipsing binaries on the basis of the presence of ellipsoidal light variations. These variations can be used to detect stellar secondaries with masses of ∼0.2 M_☉orbiting Sun-like stars at a photometric accuracy level that has already been achieved in transit surveys. By removing candidates exhibiting this effect, it is possible to greatly reduce the number of objects requiring spectroscopic follow-up with large telescopes. Unlike the usual candidate selection methods, which are based primarily on the estimated radius of the orbiting object, this technique is not biased against bona fide planets and brown dwarfs with large radii because the amplitude of the effect depends on the transiting object's mass and orbital distance. In many binary systems where a candidate planetary transit is actually due to the partial eclipse of two normal stars, the presence of flux variations due to the gravity darkening effect will show the true nature of these systems.

We show that many of the recent OGLE III photometric transit candidates exhibit the presence of significant variations in their light curves and are likely to be due to stellar secondaries. We find that the light curves of white dwarf transits generally will not mimic those of small planets because of significant gravitationally induced flux variations. We discuss the relative merits of methods used to detect transit candidates that are due to stellar blends rather than planets. We outline how photometric observations taken in two bands can be used to detect the presence of stellar blends.