MIGRATION THEN ASSEMBLY: FORMATION OF NEPTUNE-MASS PLANETS INSIDE 1 AU

We demonstrate that the observed distribution of hot Neptune/super-Earth systems is well reproduced by a model in which planet assembly occurs in situ, with no significant migration post-assembly. This is achieved only if the amount of mass in rocky material is similar to 50-100M(circle times) interior to 1 AU. Such a reservoir of material implies that significant radial migration of solid material takes place, and that it occurs before the stage of final planet assembly. The model not only reproduces the general distribution of mass versus period but also the detailed statistics of multiple planet systems in the sample. We furthermore demonstrate that cores of this size are also likely to meet the criterion to gravitationally capture gas from the nebula, although accretion is rapidly limited by the opening of gaps in the gas disk. If the mass growth is limited by this tidal truncation, then the scenario sketched here naturally produces Neptune-mass objects with substantial components of both rock and gas, as is observed. The quantitative expectations of this scenario are that most planets in the hot Neptune/super-Earth class inhabit multiple-planet systems, with characteristic orbital spacings. The model also provides a natural division into gas-rich (hot Neptune) and gas-poor (super-Earth) classes at fixed period. The dividing mass ranges from similar to 3M(circle times) at 10 day orbital periods to similar to 10M(circle times) at 100 day orbital periods. For orbital periods <10 days, the division is less clear because a gas atmosphere may be significantly eroded by stellar radiation.