LAM - Laboratoire d’Astrophysique de Marseille

Over the last 10 billion years the typical activity level of star-forming galaxies has decreased by more than an order of magnitude. Observations reveal that, at both low and and high redshift, there is a population of ‘starburst’ galaxies that are significantly more active than the bulk of the rest of the population. Despite being comparatively rare, these starbursts are among the most popular targets for detailed studies of astrophysical processes, by virtue of being bright and thus readily observable. But what does it take to produce a starburst event during the peak epoch of galaxy formation at z 2, when even the average star-formation rate of the star-forming galaxy population was as high as locally observed only for the most extreme starbursting systems ?
In my talk I will present new measurements of the composition of the starburst population at z<2, in terms of its split into ‘normal’ and interacting or merging galaxies. I will also review how a high star-formation efficiency - one of the trademarks of starburst events - is linked to galaxy environment based on the increasing body of literature on the interstellar medium in high-z galaxy clusters. Finally, I will discuss how measurements of host galaxy star-formation efficiency can provide clues on the interplay between starburst and AGN activity.

In cosmology bias is used in two different contexts. In one case it refers to systematic errors in a measurement or a determination of model’s parameters. In the other case it refers to the relation between the clustering of mass and that of observable tracers such as galaxies.
I will touch upon both aspects. I will highlight the importance of considering and modelling the effect of systematic errors and I will present a model of the halo bias in massive neutrinos cosmologies.