Using a global 3D, fully self-consistent, multi-fluid hydrodynamic model, we simulate the escaping upper atmosphere of the warm Neptune GJ436b, driven by the stellar XUV radiation impact and gravitational forces and interacting with the stellar wind. Under the typical parameters of XUV flux and stellar wind plasma expected for GJ436, we calculate in-transit absorption in Lyα and find that it is produced mostly by Energetic Neutral Atoms outside of the planetary Roche lobe, due to the resonant thermal line broadening. At the same time, the influence of radiation pressure has been shown to be insignificant. The modelled absorption is in good agreement with the observations and reveals such features as strong asymmetry between blue and red wings of the absorbed Lyα line profile, deep transit depth in the high velocity blue part of the line reaching more than 70%, and the timing of early ingress. On the other hand, the model produces significantly deeper and longer egress than in observations, indicating that there might be other processes and factors, still not accounted, that affect the interaction between the planetary escaping material and the stellar wind. At the same time, it is possible that the observational data, collected in different measurement campaigns, are affected by strong variations of the stellar wind parameters between the visits, and therefore, they cannot be reproduced altogether with the single set of model parameters.Keywords: hydrodynamicsplasmasplanets and satellites: individual: exoplanetsplanets and satellites: physical evolutionplanets and satellites: atmosphereplanet-star interactions 2004, García Muñoz 2007, Koskinen et al. 2007) clarified the basic physics of the escaping upper atmosphere in the form of planetary wind (further PW), which includes the XUV heating, hydrogen plasma photo-chemistry, radiation cooling, gravitational and thermal pressure forces. They helped to explain some of the in-transit spectral observations by the presence of an expanded partially ionized upper atmospheres, which fill the Roche lobes of hot giant exoplanets, such as HD209458b and HD189733b (Ben-Jaffel 2007, Ben-Jaffel & Sona Hosseini 2010, Koskinen et al. 2007. These expanding atmospheres were shown to be sufficiently dense to produce the absorption in Lyα due to natural line broadening mechanism.However, the detection of absorption in the resonant lines of heavy elements such as OI, CII, and SiIII , Linsky et al. 2010, has shown that the absorbing material of planetary origin far beyond the Roche lobe has to be considered as well (Ben-Jaffel & Sona Hosseini 2010, Shaikhislamov et al. 2018a. The presence of a huge hydrogen corona also is a prerequisite for the explanation of strong in-transit Lyα absorption of GJ436b. By this, there is another crucial factor, besides of the Roche lobe effect, which has to be properly taken into account in the modeling of large-scale plasma dynamics around the close-orbit exoplanetsthe stellar wind (further SW) plasma. Self-consistent description of the escaping multi-co...

The absorption of stellar radiation observed by HD 209458b in the resonant lines of O i and C ii has not yet been satisfactorily explained. We apply a 2D hydrodynamic multi-fluid model that self-consistently describes the expanding planetary wind, driven by stellar XUV radiation and influenced by tidal forces and the surrounding stellar wind. According to this model, HD 209458b has a hydrogen-dominated plasmasphere, expanding beyond the Roche lobe, in the form of two supersonic streams that propagate toward and away from the star. The species heavier than hydrogen and helium are dragged in the escaping material streams and accelerated up to 50 km s−1. Our simulations show that, assuming solar abundances, O i and C ii produce absorption due to the Doppler resonance mechanism at the level of 6%–10%, which is consistent with the observations. Most of this absorption takes place in the streams. The transit depth in the O i and C ii lines is unaffected by the stellar wind, unless it is strong enough to form a compact bowshock around the planet and able to redirect all the escaping material to the tail. In this case, the absorption profile becomes asymmetric due to the prominent blueshifted attenuation. Thus, the spectroscopic measurements enable probing of the planetary wind character, as well as the strength of the stellar wind. The computed absorption at wavelengths of the Si iii, Mg i, and Mg ii lines at solar abundances appears to be much stronger, compared to the observations. This possibly indicates that Si and Mg may be under-abundant in the upper atmosphere of HD 209458b.

Context. π Men c is the first planet to have been discovered by the Transiting Exoplanet Survey Satellite. It orbits a bright, nearby star and has a relatively low average density, making it an excellent target for atmospheric characterisation. The existing planetary upper atmosphere models of π Men c predict significant atmospheric escape, but Lyα transit observations indicate the non-detection of hydrogen escaping from the planet. Aims. Our study is aimed at constraining the conditions of the wind and high-energy emission of the host star and reproducing the non-detection of Lyα planetary absorption. Methods. We modelled the escaping planetary atmosphere, the stellar wind, and their interaction employing a multi-fluid, three-dimensional hydrodynamic code. We assumed a planetary atmosphere composed of hydrogen and helium. We ran models varying the stellar high-energy emission and stellar mass-loss rate, and, for each case, we further computed the Lyα synthetic planetary atmospheric absorption and compared it with the observations. Results. We find that a non-detection of Lyα in absorption employing the stellar high-energy emission estimated from far-ultraviolet and X-ray data requires a stellar wind with a stellar mass-loss rate about six times lower than solar. This result is a consequence of the fact that, for π Men c, detectable Lyα absorption can be caused exclusively by energetic neutral atoms, which become more abundant with increasing velocity or density of the stellar wind. By considering, instead, that the star has a solar-like wind, the non-detection requires a stellar ionising radiation about four times higher than estimated. The reason for this is that despite the fact that a stronger stellar high-energy emission ionises hydrogen more rapidly, it also increases the upper atmosphere heating and expansion, pushing the interaction region with the stellar wind farther away from the planet, where the planet atmospheric density that remains neutral becomes smaller and the production of energetic neutral atoms less efficient. Conclusions. Comparing the results of our grid of models with what is expected and estimated for the stellar wind and high-energy emission, respectively, we support the idea that it is likely that the atmosphere of π Men c is not hydrogen-dominated. Therefore, future observations should focus on the search for planetary atmospheric absorption at the position of lines of heavier elements, such as He, C, and O.