The mass determination challenge for exoplanetary science

The mass of an exoplanet is a key parameter for the characterisation of the internal structure of a planet, as well as the study of the formation and the evolution of the planet, and its atmosphere. The radial velocity technique allows for measuring the planetary mass from the radial velocity variation of its parent star. However, limitations in the property determination of exoplanets, particularly in their masses, can arise from various sources especially from astrophysical noise due to stellar variability, caused by magnetic activity, which affects the detection and characterisation of exoplanets. This PhD thesis aims to understand the impact of our knowledge of the planetary mass in the planetary atmospheric characterisation and to reduce the sources of uncertainty by a deep study of the stellar activity and by developing new techniques for stellar variability filtering. To this end, I analysed the impact of the planetary mass uncertainties of atmospheric retrievals of multiple targets from the mission reference sample of Ariel, the forthcoming ESA M4 mission aimed at studying planetary atmospheres. I simulated different spectra as observed by Ariel, assuming a primordial or secondary atmosphere of hot Jupiters, and sub-Neptunes or super-Earths, respectively, under different cloudy configurations. I estimated both the accuracy and precision necessary for each analysed target, testing also the capability of retrieval in the case of incorrect mass estimation. I verified that one of the most crucial issues is the presence of high-altitude clouds, in particular in the secondary atmosphere cases. For this reason, I tested the capability to retrieve the cloudy configuration or the presence of a secondary atmosphere during the first tier of the Ariel mission, to take an informed decision if including the planet in the Tier-2 sample. In the second part of this thesis, I described SpotCCF, a photospheric stellar model that I developed to optimise the radial velocity extraction in fast-rotating stars. This model, based on the cross-correlation function technique, takes into account the contribution of stellar activity by considering the presence of multiple spots on the stellar surface that caused deformation of the profile of the cross-correlation function. I applied this model to the HARPS-N observations of V1298 Tau, a very active K1 star, which shows strongly deformed cross-correlation function (CCF) profiles. The SpotCCF model is also able to give information about the spot configuration (latitude, longitude and area covered by the spot). In the end, I also focused my study on understanding stellar activity in M dwarfs, which is crucial for improving our understanding of the physics of stellar atmospheres and for planet search programs. Specifically, I analysed HARPS and HARPS-N observation of AD Leonis, measuring the line profiles and intensities of sensitive activity indicators, and evaluating the correlations between them. Globally, the PhD thesis highlights the importance of planetary mass characterisation and the complexity of their determination due to the effects of stellar variability. In the context of the Ariel mission, it highlights the importance of a detailed and individual analysis of each target of the mission reference sample, to be able to accurately select the Tier-2 targets and characterise their planetary atmosphere, and represents a step forward towards the preparation of the ESA M4 Ariel mission. It also shows how this work cannot be disentangled from a detailed study of the stellar variability that is crucial in the determination of the planetary mass, both in its accuracy and precision.