DO GIANT PLANETS SURVIVE TYPE II MIGRATION?

Abstract

Planetary migration is one of the most serious problems to systematically understand the observations of exoplanets. We clarify that the theoretically predicted type II migration is too fast, as well as type I migration, by developing detailed analytical arguments in which the timescale of type II migration is compared with the disk lifetime. In the disk-dominated regime, the type II migration timescale is characterized by a {\it local} viscous diffusion timescale, while the disk lifetime characterized by a {\it global} diffusion timescale that is much longer than the local one. Even in the planet-dominated regime where the inertia of the planet mass reduces the migration speed, the timescale is still shorter than the disk lifetime except in the final disk evolution stage where the total disk mass decays below the planet mass. This suggests that most giant planets plunge into the central stars within the disk lifetime, and it contradicts the exoplanet observations that gas giants are piled up at $r \ga 1$AU. We examine additional processes that may arise in protoplanetary disks: dead zones, photoevaporation of gas, and gas flow across a gap formed by a type II migrator. Although they make the type II migration timescale closer to the disk lifetime, we show that none of them can act as an effective barrier for rapid type II migration with the current knowledge of these processes. We point out that gas flow across a gap and the fraction of the flow accreted onto the planets are uncertain and they may have a potential to solve the problem. Much more detailed investigation for each process may be needed to explain the observed distribution of gas giants in extrasolar planetary systems.