Colliding planetary and stellar winds: charge exchange and transit spectroscopy in neutral hydrogen

Abstract

When transiting their host stars, hot Jupiters absorb about 10% of the light in the wings of the stellar Lyman-alpha emission line. The absorption occurs at wavelengths Doppler-shifted from line center by +/- 100 km/s - larger than the thermal speeds with which partially neutral, 10E4 K hydrogen escapes from hot Jupiter atmospheres. It has been proposed that the absorption arises from 10E6 K hydrogen from the host stellar wind, made momentarily neutral by charge exchange with planetary H I. The +/-100 km/s velocities would then be attributed to the typical velocity dispersions of protons in the stellar wind - as inferred from spacecraft measurements of the Solar wind. To test this proposal, we perform 2D hydrodynamic simulations of colliding hot Jupiter and stellar winds, augmented by a chemistry module to compute the amount of hot neutral hydrogen produced by charge exchange. The Kelvin-Helmholtz instability mixes the two winds; in the mixing layer, charge exchange reactions establish, within tens of seconds. Enough hot neutral hydrogen is generated to reproduce the transit observations, and the amount of absorption converges with both spatial resolution and time. Our calculations support the idea that charge transfer between colliding winds correctly explains Lyman-alpha transit observations - modulo the effects of magnetic fields, which we do not model but which may suppress mixing. Other neglected effects include, in order of decreasing importance, rotational forces related to orbital motion, gravity, and stellar radiation pressure; we discuss quantitatively the errors introduced by our approximations. How hot stellar hydrogen cools when it collides with cold planetary hydrogen is also considered; a more careful treatment of how the mixing layer thermally equilibrates might explain the recent detection of Balmer H-alpha absorption in transiting hot Jupiters.