Science

My research aims at understanding the formation and evolution of galaxies, from the formation of primordial haloes juste after the big bang, all the way to the galaxies we see today, including of course our own, the Milky Way. To this end I use a variety of tools. My main focus is on numerical simulations and semi-analitycal models, and my main collaborator is Dominique Aubert.

First stars and reionization

As a member of the LIDAU collaboration (Light In the Dark Ages of the Universe), I study the formation of the first stars and their main consequences: the reionization of the Universe and the photo-evaporation of the gas of low-mass galaxies. I run numerical simulations of the reionisation of the Universe, using a radiative transfer code named ATON. ATON's ability to run on GPUs, makes it possible to run massive experiments very efficiently. The figure below shows a reionisation map of the early Milky Way and its neighborhood (blue reionizes earkly, pink reionizes late).

Gas physics in galaxy formation

The advent of large cosmological simulations of galaxy formation with hydrodynamics, such as the HORIZON-MareNostrum simulation, has shed new light on the physics of galaxy formation. One of the most striking realizations that has emerged from these works is the existence of cold gas flows feeding directly galaxy disks, while it was previously believed that they formed out of gas cooling from their own hot halo. This short movie is based on the HORIZON-MareNostrum simulation and shows the formation and evolution of a galaxy hosted by a 10^12 Msun dark matter halo. Colors code: red: density, green: metals, blue: temperature. Note the ubiquity and persistence of the cold dense flows (red filaments).

STECKMAP

I have extensively studied the stellar populations of galaxies using spectroscopy and stellar population models. I wrote a code called STECKMAP (STEllar Content and Kinematics via Maximum A Posteriori likelihood), which constrains the stellar content of galaxies from their absorption line spectra. The code is public and free to use. It can be found here, along with more information on this research activity.