Vortices on Earth are known to concentrate solid particle in the eye of the vortex. A similar process is proposed to concentrate solids in protoplanetary disks and participate to planet formation. One of the most challenging step in planet formation theory is the one explaining the formation of planetesimals of kilometre size. Vortex scenario is in vogue since the asymmetries observed in protoplanetary disks by the Atacama Large Millimeter Array (ALMA) are proposed to be observational evidences of the presence of Rossby vortices in the disks. In this talk, I will introduce the vortex scenario for planetesimals formation and discuss the formation of large scale vortices in protoplanetary disks by the Rossby wave instability. I will present an analytical study and global numerical simulations of this instability leading to the formation of Rossby vortices. I will then detail the 3D structure of these vortices and its influence on solid concentration. I will conclude by comparing numerical study of solid concentration in vortices with ALMA observations.

Le jeudi 10 octobre 2013 de 11:00 à 12:00 Vortices which have long been studied in planetary atmospheres can also play an important role in the dynamics of astrophysical accretion disks. It was shown that vorticity waves can destabilize the entire disk and form large scale vortices. Such an instability is proposed to explain the large scale oscillation observed in black hole accretion disks but also to help planet formation in protoplanetary disks. After a presentation of the mechanism of this instability and the physics of the vorticity waves (Rossby waves), I will discuss the different astrophysical implications. I will then present global numerical simulations and discuss how they modified our understanding of the structure of the vortices. Finally, bi-fluid simulations will be presented giving new clues for the planet formation theory that have started to be compared with ALMA observations.

What can be proved about systems of polynomials in many variables (possibly with some parameters)? This talk will survey techniques for computing the dimension of a variety, and enumeration of the solutions if they are zero-dimensional. The main subjects will be tropical geometry, Groebner bases, amoeba, and the mixed volume. Practical implementations and strategies will be emphasized over theoretical results. Some examples will be highlighted from celestial mechanics and classical mathematical physics.

Artificial Neural Networks (ANNs) were developed to model biological information processing systems. Like biological neural networks, ANNs can learn and are often considered to be a form of artificial intelligence. Subsequent work with ANNs demonstrated that they are excellent at classifying noisy data, interpolating in high-order parameter space, and yield classifications that are equivalent to Bayesian probabilities. I will present an ANN study of instrumental trade-offs for possible Terrestrial Planet Finder instruments and a new ANN approach to fitting exoplanet spectroscopy to model planetary atmospheres.