Evaporation of Low-Mass Planet Atmospheres: Multidimensional Hydrodynamics with Consistent Thermochemistry

Evaporation of Low-Mass Planet Atmospheres: Multidimensional Hydrodynamics with Consistent Thermochemistry

Authors:

Wang et al

Abstract:

Direct and statistical observational evidences suggest that photoevaporation is important in eroding the atmosphere of sub-Neptune planets. We construct full hydrodynamic simulations, coupled with consistent thermochemistry and ray-tracing radiative transfer, to understand the physics of atmospheric photoevaporation caused by high energy photons from the host star. We identify a region on the parameter space where a hydrostatic atmosphere cannot be balanced by any plausible interplanetary pressure, so that the atmosphere is particularly susceptible to loss by Parker wind. This region may lead an absence of rich atmosphere (substantially H/He) for planets with low mass (M ~ 3 M_earth). Improving on previous works, our simulations include detailed microphysics and a self-consistent thermochemical network. Full numerical simulations of photoevaporative outflows shows a typical outflow speed ~ 30 km/s and Mdot ~ 4e-10 M_earth/yr for a 5 M_earth fiducial model rocky-core planet with 1e-2 of its mass in the atmosphere. Supersonic outflows are not quenched by stellar wind ram pressure (up to 5 times the total pressure at transonic points of the fiducial model). The outflows modulated by stellar wind are collimated towards the night side of the planet, while the mass loss rate is only ~ 25% lower than the fiducial model. By exploring the parameter space, we find that EUV photoionization is most important in launching photoevaporative wind. Other energetic radiation, including X-ray, are of secondary importance. The leading cooling mechanism is ro-vibrational molecular cooling and adiabatic expansion rather than recombination or Ly alpha cooling. The wind speed is considerably higher than the escape velocity at the wind base in most cases, hence the mass loss rate is proportional to the second power of the EUV photosphere size R_euv, instead of the third, as suggested by previous works...