2018A&A...616A.151C

We used the HARPS-North high resolution spectrograph (R=115000) at Telescopio Nazionale Galileo (TNG) to observe one transit of the highly irradiated planet MASCARA-2b/KELT-20b. Using only one transit observation, we are able to clearly resolve the spectral features of the atomic sodium (NaI) doublet and the Hα line in its atmosphere, which are corroborated with the transmission calculated from their respective transmission light curves (TLC). In particular, we resolve two spectral features centered on the NaI doublet position with an averaged absorption depth of 0.17±0.03% for a 0.75Å bandwidth with line contrasts of 0.44±0.11% (D₂) and 0.37±0.08% (D₁). The NaI TLC have also been computed, showing a large Rossiter-McLaughlin (RM) effect, which has a 0.20±0.05% NaI transit absorption for a 0.75Å passband that is consistent with the absorption depth value measured from the final transmission spectrum. We observe a second feature centered on the Hα line with 0.6±0.1% contrast and an absorption depth of 0.59±0.08% for a 0.75Å passband that has consistent absorptions in its TLC, which corresponds to an effective radius of R_λ/R_P=1.20±0.04. While the signal-to-noise ratio (S/N) of the final transmission spectrum is not sufficient to adjust different temperature profiles to the lines, we find that higher temperatures than the equilibrium (T_eq=2260±50K) are needed to explain the lines contrast. Particularly, we find that the Na I lines core require a temperature of T=4210±180K and that Hα requires a temperature of T=4330±520K. MASCARA-2b, like other planets orbiting A-type stars, receives a large amount of UV energy from its host star. This energy excites the atomic hydrogen and produces Hα absorption, leading to the expansion and abrasion of the atmosphere. The study of other Balmer lines in the transmission spectrum would allow the determination of the atmospheric temperature profile and the calculation of the lifetime of the atmosphere with escape rate measurements. In the case of MASCARA-2b, residual features are observed in the Hβ and Hγ lines, but they are not statistically significant. More transit observations are needed to confirm our findings in Na I and Hα and to build up enough S/N to explore the presence of Hβ and Hγ planetary absorptions.