ar X iv : a st ro - p h / 05 02 36 3 v 1 1 7 Fe b 20 05 Microlensing optical depth toward the Galactic Bulge using bright sources from OGLE - II

We present a measurement of the microlensing optical depth in the direction of the Galactic Bulge (GB) based on the OGLE-II survey. Using a sample of 33 events we obtained a value τ = 1.96 +0.41 −0.34 × 10 −6 along the effective line of sight (l, b) = (1. • 16, −2. • 75). Events were selected in a timescale range 1 < t E < 400 d. We considered only bright sources in the extended Red Clump Giant (RCG) region of the Color Magnitude Diagram (CMD), i.e. above the extinction corrected I-band magnitude ∼ 15.5. A gradient along the Galactic latitude b was clearly detected: τ = [(4.73 ± 1.74) + (1.02 ± 0.58) × b] × 10 −6. The present result, adjusted for the gradient, is consistent with published values based on RCG sources and recent theoretical predictions. Our analysis utilizes all 4 years of the monitoring data for the 20 bulge dominated OGLE-II fields. The fields were selected to be nearly co-located and cover the area of about 5 deg 2 between 0 • < l < 3 • and −4 • < b < −2 •. Contrary to previous analyses, we do not assume that the apparent Red Clump sources are free of blending. Photometric quality delivered by Difference Image Analysis (DIA) allows us to reject the vast majority of strong blends. About ∼ 50% of OGLE-II events which appear to have RCG sources are actually due to much fainter stars blended with a bright companion. This result contradicts a widespread belief that the number of spurious Red Clump events induced by blending is insignificant. Our conclusion is fully supported by extensive Monte-Carlo simulations of stellar crowding in the OGLE-II images. We – 2 – show explicitly that neglecting blends leads to similar optical depth estimates through partial cancellation of contributions from higher detection efficiency, underestimated timescales and larger number of selected events. Despite this lucky coincidence, strong source confusion effects in typical ground based GB surveys will still bias timescale distributions and event rates, even for relatively bright stars.