Gravitational Lensing by Spiral Galaxies

We study gravitational lensing by spiral galaxies, using realistic models consisting of halo, disk, and bulge components combined to produce a flat rotation curve. Proper dynamical normalization of the models is critical because a disk has less (projected) mass than does a spherical halo with the same rotation curve—a face-on Mestel disk has a lensing cross section only 41% as large as that of a comparable singular isothermal sphere. Lensing cross sections are sensitive to inclination and are dominated by edge-on galaxies that, if very flattened, produce a large number of lenses with an unobserved image geometry consisting of 2 or 3 images off to one side of the galaxy center and straddling the projected disk. The absence of this "disk" image geometry among known lenses suggests that lens galaxies cannot be highly flattened. When averaged over inclination, disk+halo models predict ≲ 10% more lenses than do pure halo models, except in cases in which the disk is unreasonably massive, so including a disk does not significantly increase the expected number of spiral galaxy lenses. Models with an exponential disk and a central bulge are sensitive to the properties of the bulge. In particular, an exponential disk model normalized to our Galaxy but without a bulge cannot produce multiple images, and including a bulge reduces the net flattening of edge-on galaxies. The dependence of the lensing properties on the masses and shapes of the halo, disk, and bulge means that a sample of spiral galaxy lenses would provide useful constraints on galactic structure.