2008ApJ...677..607B

Both core accretion and disk instability appear to be required as formation mechanisms in order to explain the entire range of giant planets found in extrasolar planetary systems. Disk instability is based on the formation of clumps in a marginally gravitationally unstable protoplanetary disk. These clumps can only be expected to contract and survive to become protoplanets if they are able to lose thermal energy through a combination of convection and radiative cooling. Here we present several new three-dimensional, radiative hydrodynamics models of self-gravitating protoplanetary disks, where radiative transfer is handled in the flux-limited diffusion approximation. We show that while the flux-limited models lead to higher midplane temperatures than in a diffusion approximation model without the flux limiter, the difference in temperatures does not appear to be sufficiently high to have any significant effect on the formation of self-gravitating clumps. Self-gravitating clumps form rapidly in the models both with and without the flux limiter. These models suggest that the reason for the different outcomes of numerical models of disk instability by different groups cannot be attributed solely to the handling of radiative transfer, but rather appears to be caused by a range of numerical effects and assumptions. Given the observational imperative to have disk instability form at least some extrasolar planets, these models imply that disk instability remains as a viable giant planet formation mechanism.