Low Microlensing Optical Depth Toward the Galactic Bar

Abstract

I make a new evaluation of the microlensing optical depth toward the Galactic bar from Difference Image Analysis (DIA) of the MACHO Collaboration. First, I present supplementary evidence that MACHO field 104 located at (l,b) = (3.11,-3.01) is anomalous in terms of the event duration distribution. I argue that both the event durations and the very high optical depth of field 104 are not representative and, therefore, exclude this field as an outlier. In addition, I eliminate field 159 at (l,b) = (6.35,-4.40) based mainly on its separate location, but also on unexplained statistical properties of the event durations. The remaining six DIA fields form a very homogeneous and spatially compact set that is very suitable for averaging. The weighting of the optical depth values for these six DIA fields results in a total optical depth tau_{tot} = 2.01^{+0.34}_{-0.32} x 10^{-6} at (l,b) = (2.22,-3.18). If a fraction of all sources, f_{disk}, assumed to be in the disk, does not contribute to microlensing, then the optical depth toward the sources in the bar is tau_{bar} = 2.23^{+0.38}_{-0.35} x 10^{-6} 0.9/(1-f_{disk}). Both tau_{tot} and tau_{bar} are substantially lower than the original estimates of Alcock et al. Most of the change in the DIA-based optical depths comes from a more appropriate statistical treatment of the results in individual fields and not from the removal of fields 104 and 159. When taken together with tau_{bar} = (1.4 +/- 0.3) x 10^{-6} at (l,b) = (3.9,-3.8) as derived from clump giants, this new result suggests that the conclusions from microlensing experiments are in reasonable agreement with the expectations from infrared-based Galactic models.