Galaxies and halosFeedback and baryon cyclesInitial conditions

Galaxies and halos

I'm very interested in the formation, and evolution, of galaxies and in particular, the co-evolution of galaxies with their hot atmospheres of gas sometimes known as the circumgalactic medium (CGM).

In general, stars make up only a small fraction of the ordinary matter that we expect to find in the universe. Thus a large fraction of ordinary matter is unaccounted. My research of the EAGLE simulations finds that a large fraction of the ordinary matter is in gaseous halos which surround galaxies. My work also focuses on making predictions for how future telescopes and instruments can detect these hot halos.

Hot halo
Left: Visible view of galaxy, Right: The X-ray view of the hot halo

Feedback and baryon cycles

Most modern hydrodynamical simulations of the universe can accurately recover many properties of the real universe. These simulations follow the evolution of galaxies and model a variety of physical processes.

However, due to the limited resolution of the simulations, many of the processes which occur in a galaxy can not be modelled correctly. Instead, the effects of the physics are included via a subgrid model. Although many distinct subgrid models have proven to be successful in modelling the effects of supernovae (exploding stars) feedback on the primary galaxy properties, it is still unclear if, and by how much, different versions of subgrid models affect other aspects of galaxies.

I’m interested in investigating the ‘secondary’ properties of galaxies; the properties that are both untuned and difficult to observe. To do this, we set up experiments where we simulate the same patch of the universe with different models.


Initial conditions

Cosmological N-body simulations proceed by evolving the forces of gravity on an ensemble on particles. In these simulations, we want to both evolve a large volume and ensure that each halo is resolved with a large number of particles.

However, this is prohibitive. Instead, techniques have been developed in which a large volume of the universe can be simulated at low resolution. Upon completion, regions of interest in the large parent volume simulation can be selected and re-simulated at a much higher resolution. These are called ‘zoom in’ simulations.

My work focusses on developing efficient pipelines to select and resimulate volumes from a parent simulation.