ARE THE GIANT PLANETS FULLY CONVECTIVE

All current models of internal structure and evolution of the giant pl anets assume a convective and adiabatic stratification. The purpose of this work is to reexamine this hypothesis using the most recent avail able opacity data. Rosseland opacities are calculated for density and temperature conditions relevant to the fluid molecular envelope of the giant planets and for the compositions inferred from atmospheric meas urements. These opacities include the absorption by hydrogen, helium, water, methane, and ammonia. The contribution of metals to the electro n abundance is also taken into account. The region investigated covers temperatures from 200 to 5000 K and densities from 10(-4) to 0.5 g cm (-3) It is shown that giant planets of pure hydrogen and helium may no t be fully convective for temperatures below similar to 4000 K. The op acity due to water and methane is very strong at low temperature level s but is insufficient to restore convection around similar to 2000 K i n Jupiter and Saturn; however, the lack of data precludes us from safe ly modeling the ammonia opacity at wavenumbers higher than 2000 cm(-1) . Our calculations also predict a radiative zone in Uranus but the unc ertainty in the value of the opacity prevents any firm conclusion. The fluid envelope of Neptune seems to be entirely convective. The possib le influence on the absorption of heavy constituents other than CH4, N K3, and H2O is discussed. (C) 1994 Academic Press, Inc.