NASA@SC19

The Universe

Simulating the Diffuse Gas Surrounding Galaxies

Overview

Galaxies grow by converting gas into stars. They acquire this gas from their diffuse environs, known as the circumgalactic medium (CGM). As stars evolve and die as explosive supernovas, they eject freshly produced heavy elements such as oxygen and iron back into the CGM, along with prodigious amounts of energy and momentum. Some of this gas will then recycle back into galaxies to fuel future generations of star formation. Thus, understanding the physics of this dynamic and structurally complex medium is crucial to understanding how galaxies form and evolve. Thanks in part to thousands of hours of observations made with NASA’s Hubble Space Telescope and other observatories, we now know the CGM to be a massive and extensive reservoir of galactic gas and heavy elements. Circumgalactic gas is generally too diffuse to be seen in emission, so rather than taking pictures of it, we observe it in absorption towards bright background light sources, akin to using cosmic lighthouses to observe this cosmic fog.

Project Details

The Figuring Out Gas & Galaxies In Enzo (FOGGIE) project uses the Enzo cosmological hydrodynamic code to model the co-evolution of galaxies and their gas, with a focus on resolving the ultra-diffuse CGM with unprecedented fidelity. We “observe” these simulations in a manner similar to how the real CGM is observed; these synthetic data help us interpret the real observations as, unlike in the real universe, we know what the underlying gas physics is in the simulations.

Results and Impact

These simulations reveal a richly structured circumgalactic medium full of churning turbulent gas, small clouds, and tenuous hot gas. Observations made with Hubble have shown that low-ionization gas, which should be relatively cool and higher-density, often has kinematic structure very similar to more highly ionized gas, which is expected to generically be hotter and lower-density. With FOGGIE’s exquisite resolution, we have started to disentangle this puzzle: sometimes kinematically coincident gas is actually co-spatial, but often it is not; rather, the observation is a chance superposition along the line of sight. By directly relating the observational signatures in the synthetic data back to their underlying physical conditions, we are learning how to better glean information about the real CGM and its relation to galaxies.

Why HPC Matters

The FOGGIE simulations could not be run without high-performance computing resources. Evolving a single galaxy and its gaseous halo at high resolution across cosmic time typically takes a few million core-hours on the Pleiades supercomputer at the NASA Advanced Supercomputing (NAS) facility; this corresponds to wall-clock times of over a year and hundreds of thousands of node-hours on the system’s 128-GB-memory Intel Xeon Broadwell and Haswell nodes. The large memory per core is critical because these kinds of simulations are particularly memory intensive: they are complex, multiphysics calculations following many fields (density, temperature, chemical composition, fluxes, and so on), combined with the tens of millions of resolution elements needed to resolve the CGM. In addition, the long-term mass storage system at the NAS facility is important for enabling analysis: each output can be ~15 gigabytes, and with 2,500 outputs per run, it’s crucial to have access to hundreds of terabytes of space with fast I/O in order to analyze the detailed time evolution of the galaxies and their gas.

What's Next

The next steps for FOGGIE are to extend the small-scale resolution of the CGM to larger volumes for more galaxies and over more of cosmic time. To do this, we are transitioning to the Enzo-E code, an updated version of Enzo that is scalable to thousands of processors.

More Information

Molly S. Peeples, Space Telescope Science Institute/Johns Hopkins University
molly@stsci.edu