Abstract :
[en] Following the advent of space-borne missions (e.g. CoRoT, Kepler), came a wealth of data of
unprecedented quality. This enabled asteroseismology to thrive and to probe the stellar structure
of a wide variety of pulsating stars. Amongst these pulsating stars is the notable category of
low-mass stars. These exhibit masses below 2.3 solar-mass , encompassing the case of our Sun.
Throughout their evolution, these stars exhibit a few interesting peculiarities. First, during the
main-sequence phase, they display a very regular pressure-mode oscillation spectrum. However,
small perturbations to that regularity may occur. Such perturbations are the result of sharp and
localised variations in the stellar structure. These create an oscillating feature, as a function of
the frequency, in the oscillation spectrum, the so-called glitches. These glitches are of particular
interest as they allow us to probe very localised regions of the stellar interior and provide diagnoses
about specific stellar features, inaccessible by other means. In main-sequence low-mass stars, we
distinguish two main causes of glitches: the helium second-ionisation zone, providing information
about the surface helium abundance, and the base of the envelope convection zone, constraining
the mixing processes in that region. The first part of my thesis was dedicated to the development
of a seismic technique, WhoSGlAd, that consistently analyses the complete oscillation spectra
of main-sequence low-mass stars and robustly retrieves the glitches signatures present in these
spectra. Special care was put in the definition of stringent seismic indicators as we decorrelated
them as much as possible. This is done thanks to a Gram-Schmidt orthonormalisation process.
The defined indicators were then used to constrain stellar models and provide a characterisation of
both the 16 Cygni system and the Kepler Legacy Sample, representing the best solar-like seismic
data currently available.
After the main-sequence phase, low-mass stars evolve on the subgiant and red-giant phases.
Their core then contracts while their envelope expands, developing a large core-envelope density
contrast. This produces the appearance of mixed-modes, presenting a twofold nature: a gravity-
dominated nature in the inner radiative regions, and a pressure-dominated nature in the outer
convective regions. These modes have the great advantage to propagate throughout most of the
stellar interior and, therefore, to probe almost the complete stellar structure. To exploit the
information these modes carry, we developed the EGGMiMoSA method. It relies on the asymptotic
expression and allows us to precisely measure seismic indicators on both subgiant and red-giant
stars. The method was applied to a grid of models extending from the subgiant phase to the
luminosity bump. The results are excellent in regard to the asymptotic values of the seismic
indicators and also qualitatively agree with observed and theoretical studies. These indicators
also allow us to efficiently infer the stellar age, mass, and radius of subgiant stars and of red-giant
stars with masses & 1.8 M . Below this threshold, we noted that the central electron degeneracy
impaired our diagnosis of the stellar age, mass, and radius in red-giants.
The combination of both methods should provide means to constrain the stellar structure of
low-mass stars from the early main-sequence phase to the late red-giant one. This is a unique
opportunity to study their structure through most of their evolution and, for example, pinpoint
missing physical processes in their modelling.