Planet atmosphere

III. Characterising the atmosphere of transiting planets in high-resolution spectroscopy

Planet atmospheres are the best laboratories to study the formation and evolution of planetary systems as well as their surface conditions an interactions with the interplanetary medium. Over the last two decades, a growing number of exoplanets have seen their atmospheres probed with low-resolution spectrographs embarked on space telescopes such as the Hubble Space Telescope and the Spitzer space telescope (see the reviews of Madhusudhan et al. 2016, 2019) . The recently-launched James Webb Space Telescope (JWST) and the forthcoming ARIEL Space Mission will mark a milestone in exoplanetary science by allowing us to probe the atmospheres of terrestrial planets with unrivalled precision. However, such spectrographs feature no more than a low-to-medium spectral resolution and are thus not able to resolve very narrow lines in the planet atmosphere spectrum. Over the past decade, ground-based high-resolution spectroscopy has emerged as a reliable method to probe key properties of exoplanet atmospheres (composition, temperature-pressure profile, atmospheric winds; see Snellen et al., 2010, de Kok et al. 2013, Brogi et al. 2016, Birkby et al. 2017, Giacobbe et al. 2021). New-generation near-infrared high-resolution spectrographs like SPIRou, GIANO-B, IGRINS or CARMENES feature a large continuous spectral band, particularly suited for probing the atmospheres of a wide range of exoplanetary atmospheres. Even though such methods work on non-transiting planets (Brogi et al. 2012, Pelletier et al. 2021), transiting planets remain the best suited for atmospheric characterisation, due to the favourable inclination of their orbit. In what follows, I will focus on transmission spectroscopy, which consists in measuring the evolution of the primary transit depth with the wavelength.

1. Transmission spectroscopy with high-resolution near-infrared spectrographs

Detection of water in the atmosphere of the hot Jupiter HD 189733 b from SPIRou observations. The Y and X axes stand respectively for the planet radial velocity amplitude and the radial velocity of the planet at the mid-transit time. The colour map indicates the value of the correlation between a 1D model containing only water lines and the data (whiter regions indicate stronger correlation with the model). The detection is found at orbital parameters consistent with the literature. In particular, we note that the planet radial velocity at mid-transit is shifted by ~-3 km/s, suggesting the presence of strong wind in the layer of planet atmosphere probed by our observations.

With its continuous spectral coverage of the Y, J, H and K bands (0.95 to 2.3 μm) and its high resolving power of ~70 000, SPIRou has the capacity to resolve a the profusion of molecular lines (notably H2O, CO, CO2, CH4) spanning the near-infrared domain. However, only a very small fraction of the stellar light crosses the planet atmosphere during the transit. As a consequence, the planetary signal to unveil is typically 10 to 100 times smaller than the photon noise, itself an order of magnitude smaller than the spectra of the star and of the Earth atmosphere (a.k.a telluric spectrum). Both contributions need to be removed in order to extract the spectrum of the exoplanet atmosphere. Throughout the transit, as the planet crosses the stellar disk, its velocity with respect to the observer varies significantly (typically of ~20 km/s for a hot Jupiter). As a result of the Doppler effect, the spectrum of the planet atmosphere will be significantly shifted during the transit. In contrast, both the Earth atmosphere and the star remains mostly immobile during the planet transit and their respective spectra will stay immobile during the transit. This is the basis of an iterative data-driven process, precisely described in Brogi et al. 2018 and in Boucher et al. 2021 (application to SPIRou data). Once this pre-processing procedure has been applied, we still need to extract the planet atmosphere spectrum from the white noise. This is done using a so-called template matching framework. We first need to generate a model for the planet atmosphere spectrum. A great variety of codes solving the radiative transfer equation for a given model of planet atmosphere are available on the EMAC NASA webpage (see also the documentation of the HITRAN line list). This model is then compared to the observations for different planet orbital properties. The maximum correlation between the model and parameters, if significantly high than the typical value of the correlation with a white noise model, is interpreted as a detection of the planet atmosphere. We coded our own version of the code, soon to be publicly available, and obtained a nearly 10 sigma detection of water in the atmosphere of the hot Jupiter HD 189733 b observed with SPIRou (5 transits; Debras, Klein et al, in prep.).

Page updated

Google Sites

Report abuse