Identifying Microlensing by Binaries

The microlensing monitoring programs have studied large numbers of standardlight curves which seem to be due to lensing by a dark point mass. Theorypredicts that many microlensing events should display significant deviationsfrom the standard form. Lens binarity in particular is expected to be common.So far, however, only a handful of light curves exhibit evidence that the lensis a binary; all of these display dramatic deviations from the standard lightcurve, exhibiting pronounced multiple peaks and/or caustic crossings.Binary-lens events in which the light curve is less dramatically perturbedshould also exist in the data set. Why, then, have we not detected them? Theanswer may lie in the fact that the perturbations, though often significant,tend to be less distinctive than those associated with caustic crossings. We present a method to determine whether a light curve is due to lensing by abinary. The method works for both gently and dramatically perturbed binary-lenslight curves. Our method identifies all degenerate solutions-- i.e., allpossible lensing events that might have given rise to the observed light curve.It also enables us to eliminate from consideration large ranges of possiblefalse positive identifications associated with light curves that might mimicmicrolensing by a binary. This method, or a generalization of it, can also beapplied to the analysis of light curves that deviate from the standardpoint-mass lens form because of astronomical effects other than lens binarity.Comment: 22 pages, 5 figures; submitted to ApJL 11 December 96; resubmitted to ApJ 7 February 199