Comment on "Simulation of a two-dimensional Rayleigh-Bénard system using the direct simulation Monte Carlo method"

In a recent paper, Watanabe, Kaburaki, and Yokokawa [Phys. Rev. E 49, 4060 (1994)] used a direct simulation Monte Carlo method to study Rayleigh-Benard convection. They reported that, using stress-free boundary conditions, the onset of convection in the simulation occurred at a Rayleigh number much larger than the critical Rayleigh number predicted by a linear stability analysis. We show that the source of the discrepancy is their omission of a temperature jump effect in the calculation of the Rayleigh number.