Absorption in the stellar Lyman-$\alpha$ (Ly-$\alpha$) line observed during the transit of the extrasolar planet HD 209458b reveals high velocity atomic hydrogen at great distances from the planet. This has been interpreted as hydrogen atoms escaping from the exosphere of the planet, possibly undergoing hydrodynamic blow-off, being accelerated by stellar radiation pressure. However, around solar system planets the production of energetic neutral atoms from charge exchange between solar wind protons and neutral hydrogen from the exospheres has been observed, and should also occur at extrasolar planets. Here we show that the measured transit-associated Ly-$\alpha$ absorption can be explained by the interaction between the exosphere of HD 209458b and the stellar wind, and that radiation pressure alone cannot explain the observation. This is the first observation of energetic neutral atoms outside the solar system. Since the stellar wind protons are the source of the observed energetic neutral atoms, this provides a completely new method of probing stellar wind conditions, and our model suggests a slow and hot stellar wind near HD 209458b at the time of the observation.

Aims: We investigate the influence of high-energy stellar radiation at close-in orbits on atmospheric mass loss during stellar evolution of a G-type star. Methods: High-energy stellar luminosity varies over a wide range for G field stars. The temporal evolution of the distribution of stellar X-ray luminosity and its influence on the evolution of close-in exoplanets is investigated. X-ray luminosity distributions from the Pleiades, the Hyades and the field are used to derive a scaling law for the evolution of the stellar X-ray luminosity distribution. A modified energy-limited escape approach is used to calculate atmospheric mass loss for a broad range of planetary parameters. Results: We show that the evolution of close-in exoplanets strongly depends on the detailed X-ray luminosity history of their host stars, which varies over several orders-of-magnitude for G stars. Stars located at the high-energy tail of the luminosity distribution can evaporate most of its planets within 0.5 AU, while for a moderate luminosity a significant fraction of planets can survive. We show the change on an initial planetary mass distribution caused by atmospheric escape.

The orbital distance at which close-in exoplanets maintain their initial mass is investigated by modelling the maximum expected thermal and nonthermal mass loss rates over several Gyr. Depending on an exosphere formation time and the evolution of the stellar X-ray and EUV flux we expect that thermal evaporation at orbital distances less than 0.05 AU may be an efficient loss process for hydrogen-rich exoplanets with masses less than 0.25 MJup. Our results indicate that nonthermal mass loss induced by Coronal Mass Ejections of the host star can significantly erode weakly magnetized short periodic gas giants. The observed exoplanets Gliese 876d at 0.0208 AU with a mass of about 0.033 MJup and 55 Cnc e at 0.045 AU with a mass of about 0.038 MJup could be strongly eroded gas giants, while HD69830b, at 0.078 AU, HD160691d at 0.09 AU and HD69830c at 0.18 AU belonged most likely since their origin to the Neptune-mass domain. The consequences for the planetary population predicted in paper I (Wuchterl et al. 2006) for CoRoTs first field are: (1) for orbital distances less than 0.05 AU (orbital periods less than days) weakly magnetized or highly irradiated gas giants may loose a large fraction of their initial mass and completely loose their gas envelopes. (2) Observed planetary mass spectra at these periods that resemble the initial ones would indicate a major effect of magnetic field protection and so far unknown thermospheric cooling processes. (3) At distances larger than 0.05 AU the impact of loss processes is minor and the observed mass spectra should be close to the initial ones.