Abstract
This thesis is part of the general study of dynamical processes involved in stars such as convection, rotation or magnetic .elds and of their nonlinear interactions. The results of numerical simulations using the 2D .nite element code STELEM and the pseudo-spectral 3D code ASH are presented.
The .rst part of this work focuses on the global modeling of the solar dynamo. Through 2D simulations using mean-.eld theory, I studied the in.uence of a complex pro.le of meridional .ow in Babcock-Leighton models. Even if the 22-yr cycle period can be reproduced by these models, the resulting butter.y diagram is a.ected to a point where it is unlikely that such multicellular meridional .ows persist for a long period of time inside the Sun, without having to reconsider the model itself. We thus show that there may be doubts about the ability of such models to reproduce the main characteristics of the solar cycle. In order to better constrain the e.ects of solar variability on the Earth climate, we present a .rst application in solar physics of sophisticated prediction methods which are used in meteorology.
I also computed the .rst 3D MHD simulations in spherical geometry of a key step in the solar dynamo : the nonlinear evolution of magnetic structures from the base of the convection zone up to the surface where they produce active regions. The global e.ects of hoop stresses and mean .ows are taken into account. Weak .elds are likely to be modulated by convective motions, thus creating favored longitudes of emergence. We show that the e.ect of convective motions on the tilt angle has to be taken into account and that Joy’s law should not be explained exclusively by the Coriolis force acting on the .ux rope. The introduction of an atmosphere in these models is a step towards a 3D global vision of our Sun.