RATES FOR COLOR-SHIFTED MICROLENSING EVENTS

If the objects responsible for gravitational microlensing of Galactic bulge stars are faint dwarfs, then blended light from the lens will di stort the shape of the microlensing light curve and shift the color of the observed star during the microlensing event. In most cases, the r esolution in current microlensing surveys is not accurate enough to ob serve this color-shift effect. However, such signatures could conceiva bly be detected with frequent follow-up observations of microlensing e vents in progress, providing the photometric errors are small enough. We calculate the expected rates for microlensing events in which the s hape distortions will be observable by such follow-up observations, as suming that the lenses are ordinary low-mass main-sequence stars in a rapidly rotating bar and in the disk. We adopt Galactic models consist ent with observed microlensing timescale distributions and consider se parately the cases of self-lensing of the bulge, lensing of the bulge by the disk, and self-lensing of the disk, further differentiating bet ween events where the source is a giant or a main-sequence star. We st udy the dependence of the rates for color-shifted microlensing events on the frequency of follow-up observations and on the precision of the photometry for a variety of wave band pairings, We find that for hour ly observations in B and K with typical photometric errors of 0.01 mag , 28% of the events where a main-sequence bulge star is lensed, and 7% of the events where the source is a bulge giant, will give rise to a measurable color shift at the 95% confidence level. For observations i n V and I, the fractions become 18% and 5%, respectively, but may be i ncreased to 40% and 13% by improved photometric accuracy and increased sampling frequency. Unlike standard achromatic lensing events, color- shifted events provide information on the lens mass, distance, and vel ocity. We outline how these parameters can be obtained, giving example s of typical errors that may arise in the calculation, and briefly dis cuss other applications of such light-curve measurements. We show that color-shifted events can be individually and/or statistically disting uished from events in which the source is blended with a binary compan ion. In particular, the observed fraction of color-shifted events as a function of event timescale will test whether the color shift is caus ed by a lens that is a low-mass main-sequence star or by a blended sta r.