Blending in Future Space-based Microlensing Surveys

Abstract

We investigate the effect of blending in future gravitational microlensing surveys by carrying out simulation of Galactic bulge microlensing events to be detected from a proposed space-based lensing survey. From this simulation, we find that the contribution of the flux from background stars to the total blended flux will be equivalent to that from the lens itself despite the greatly improved resolution from space observations, implying that characterizing lenses from the analysis of the blended flux would not be easy. As a method to isolate events for which most of the blended flux is attributable to the lens, we propose to use astrometric information of source star image centroid motion. For the sample of events obtained by imposing a criterion that the centroid shift should be less than three times of the astrometric uncertainty among the events for which blending is noticed with blended light fractions $f_{\rm B}>0.2$, we estimate that the contamination of the blended flux by background stars will be less than 20% for most ($\sim 90%$) of the sample events. The expected rate of these events is $\gtrsim 700$ events/yr, which is large enough for the statistical analysis of the lens populations.