Observational Gravitational Microlensing

Abstract

Gravitational microlensing is name given to the gravitational deflection of light emitted by a luminous source object by a massive 'lens' object which lies close to the observer-source line-of-sight. This results in an apparent time-dependent brightening of the source, although Such events are very rare. A Japanese/New Zealand collaboration, called the MOA Project, was set up in 1995 with the primary aim of detecting and observing gravitational microlensing events, particularly events which may reveal the presence of extra-solar planetary systems. The research contained within this thesis is concerned with the development of a software system, called autophot, for automatically analyzing large numbers of astronomical digital images, particularly those produced by the MOA microlensing survey. This required the author to develop algorithms and corresponding software implementations for: frame registration; overlapping mosaic coordinate system construction; fast photometric zeropoint estimation of inhomogeneous ensembles of stars; a multi-processor scheduler with the unique ability to use both the DoPHOT and DAOphot packages; and a very high speed object database. This system was designed to have performance which scales well when the number of observed objects is in the millions, as well as robustness, portability, and flexibility. As a result of the system's portability and flexibility components have been adopted by other astronomers in New Zealand, Tokyo, Rome, Nottingham, St. Andrews, Texas, and Princeton. The results of running autophot on a set of images of the pulsating white dwarf GW Librae are presented here. MOA adopted a set of custom ultra-broad band photometric filters in order to optimise the light collecting ability of their telescope. The author has compared the MOA filters with the commonly used Johnson filter set using observations of the cluster M41, the globular clusters 47 Tucanae and NGC3201, as well as numerical simulations of the optical system. The detection efficiency and effect of the atmospheric extinction on measurements made with the MOA observing system are also investigated. The spatial uniformity of the CCD camera used by MOA is also characterised. Several microlensing events observed by the MOA group are presented. One microlensing event is used to compare the speed, completeness, and accuracy of the different measurement packages automated by autophot. The microlensing event 95-BLG-30 was processed with autophot and shows finite-source microlensing effects. The the radius of the source star in this event was estimated as 0.08 times the Einstein ring radius of the lens, consistent with the radius of 0.07 Einstein ring radii from an earlier analysis by the author which was published as part of an international collaboration in the Astrophysical Journal. A high-magnification event, 98-BLG-35 was analysed with autophot producing an improved lightcurve over that originally published by the MOA collaboration, and only slightly poorer than the lightcurve results produced by the newly developed image substraction technique. Small deviations from single lens behaviour provide weak evidence for the presence of a planet around the lens star. The flexibility of autophot will allow it to incorporate its own version of image substraction in the future.