Investigating the Andromeda stream - II. Orbital fits and properties of the progenitor

We construct test-particle orbits and simple N-body models that match the properties of the giant stellar stream observed to the south of M31, using the model of M31's potential derived in the companion paper by Geehan et al. We introduce a simple approximation to account for the difference in position between the stream and the orbit of the progenitor; this significantly affects the best-fitting orbits. The progenitor orbits we derive have orbital apocentre similar to 60 kpc and pericentre similar to 3 kpc, though these quantities vary somewhat with the current orbital phase of the progenitor which is as yet unknown. Our best combined fit to the stream and galaxy properties implies a mass within 125 kpc of M31 of (7.4 +/- 1.2) x 10(11)M circle dot. Based on its length, width, luminosity, and velocity dispersion, we conclude that the stream originates from a progenitor satellite with mass M-s similar to 10(9)M circle dot, and at most modest amounts of dark matter; the estimate of Ms is again correlated with the phase of the progenitor. M31 displays a large number of faint features in its inner halo which may be progenitors or continuations of the stream. While the orbital fits are not constrained enough for us to conclusively identify the progenitor, we can identify several plausible candidates, of which a feature in the planetary nebula distribution found by Merrett et al. is the most plausible, and rule out several others. We make predictions for the kinematic properties of the successful candidates. These may aid in observational identification of the progenitor object, which would greatly constrain the allowed models of the stream.