Massive planet migration: Theoretical predictions and comparison with observations

We quantify the utility of large radial velocity surveys for constraining theoretical models of type II migration and protoplanetary disk physics. We describe a theoretical model for the expected radial distribution of extrasolar planets that combines an analytic description of migration with an empirically calibrated disk model. The disk model includes viscous evolution and mass loss via photoevaporation. Comparing the predicted distribution to a uniformly selected subsample of planets from the Lick, Keck, and AAT planet search programs, we find that a simple model in which planets formin the outer disk at a uniform rate, migrate inward according to a standard type II prescription, and become stranded when the gas disk is dispersed is consistent with the radial distribution of planets for orbital radii in the range 0.1 AU <= a < 2.5 AU and planet masses M-p > 1.65 M-J. Some variant models are disfavored by existing data, but the significance is limited (similar to 95%)due to the small sample of planets suitable for statistical analysis. We show that the favored model predicts that the planetary mass function should be almost independent of orbital radius at distances where migration dominates the massive planet population. We also study how the radial distribution of planets depends on the adopted disk model. We find that the distribution can constrain not only changes in the power-law index of the disk viscosity, but also sharp jumps in the efficiency of angular momentum transport that might occur at small radii.