X-RAY IRRADIATED PROTOPLANETARY DISK ATMOSPHERES. II. PREDICTIONS FROM MODELS IN HYDROSTATIC EQUILIBRIUM

We present new models for the X-ray photoevaporation of circumstellar disks which suggest that the resulting mass loss (occurring mainly over the radial range 10-40 AU) may be the dominant dispersal mechanism for gas around low-mass pre-main-sequence stars, contrary to the conclusions of previous workers. Our models combine use of the MOCASSIN Monte Carlo radiative transfer code and a self-consistent solution of the hydrostatic structure of the irradiated disk. We estimate the resulting photoevaporation rates assuming sonic outflow at the surface where the gas temperature equals the local escape temperature and derive mass-loss rates of similar to 10(-9) M-circle dot yr(-1), typically a factor of 2-10 times lower than the corresponding rates in our previous work where we did not adjust the density structure of the irradiated disk. The somewhat lower rates, and the fact that mass loss is concentrated toward slightly smaller radii, result from the puffing up of the heated disk at a few AU which partially screens the disk at tens of AU. Our mass-loss fluxes agree with those of Alexander et al. but we differ from Alexander et al. in our assessment of the overall significance of X-ray photoevaporation, given the large disk radii (and hence emitting area) associated with X-ray-driven winds. Gorti & Hollenbach, on the other hand, predict considerably lower mass-loss fluxes than either Alexander et al. or ourselves and we discuss possible reasons for this difference. We highlight the fact that X-ray photoevaporation has two generic advantages for disk dispersal compared with photoevaporation by extreme ultraviolet (EUV) photons that are only modestly beyond the Lyman limit: the demonstrably large X-ray fluxes of young stars even after they have lost their disks and the fact that X-rays are effective at penetrating much larger columns of material close to the star. We however stress that our X-ray-driven mass-loss rates are considerably more uncertain than the corresponding rates for EUV photoevaporation (around 10(-10) M-circle dot yr(-1)) and that this situation will need to be remedied through future radiation hydrodynamical simulations.