Petascale Adaptive Mesh Simulations of Milky Way-type Galaxies and Their Environments

Brian O'Shea, Michigan State University

Usage Details

Michael Norman, Brian O'Shea, David Collins, Britton Smith, John Wise, Cameron Hummels, Jason Tumlinson, Brian Crosby, Devin Silvia, Pengfei Chen, Claire Kopenhafer, Thomas Bolden, Austin Gilbert, Lauren Corlies, Molly Peeples, Kye Stalpes, Philipp Grete, Forrest Glines, David Crowe, Justin Grace, Bryan Brzycki

This project seeks answers to several pressing questions about the formation and evolution of galaxies. It does so by using the Blue Waters supercomputer to perform a suite of sophisticated supercomputer simulations. The investigators will address such questions as: (i) How did the earliest progenitors of the Milky Way galaxy form, and where can we find their stellar remnants today? (ii) How does the ionizing radiation produced by massive stars escape from galaxies, and how does it affect the properties of neighboring galaxies? (iii) How does the gas that is critical for star formation get from the cosmic web into the central regions of galaxies, and how is gas returned to the intergalactic medium? (iv) How are magnetic fields seeded and amplified in galaxies, and how are they ejected into (or amplified in) the intergalactic medium?

The team includes experts in astrophysics as well as in high performance computing, and is united in the use of a sophisticated numerical tool (the Enzo AMR code) that has already demonstrated its performance on Blue Waters. The team will work with observational astronomer collaborators to apply these simulations to the interpretation of measurements of both local and distant galaxies from current astronomical surveys, and to motivate future observations by the Large Synoptic Survey Telescope and the James Webb Space Telescope.

The proposed work promises to have significant impact on scientists in training, who will learn to use cutting-edge numerical tools at the largest possible scale. The project will involve undergraduate students at Michigan State University (through MSU's REU program, which targets women and under-represented minorities) and postdoctoral researchers in the research efforts. Scientific results from this program will be visualized by staff at the National Center for Supercomputing Applications, and will be disseminated to the public via pre-existing collaborations with planetaria and museums, and via the Internet. In addition, these visualizations will be used as part of outreach talks given by members of this project. Finally, the simulation data produced as a result of this project will be used in computational science courses at Michigan State University, where it will be used to train students in scientific visualization and data analysis techniques. The resulting curricular materials will be made available to the public via the World Wide Web.

The specific research methods used in this project include the creation of an extensive library of simulated Milky Way-like galaxies and their environments that can be used to explore a wide range of observable astrophysical phenomena. This will be the first study to perform cosmological simulations of galaxy formation and evolution that include self-consistent treatments of radiation transport and/or magnetohydrodynamics for a statistically significant number of galaxies, and to apply these calculations to the interpretation of recent observations relating to the intergalactic and circumgalactic medium, galactic and extragalactic magnetic fields, and high redshift galaxy formation. Furthermore, the simulation data produced during the course of this project, as well as a wide range of data products, will be made publicly available via the nascent National Data Service. This data will be usable by the astrophysical research community, and will enable researchers to address a much broader range of questions regarding galaxy formation and evolution than can be done as a part of this project alone, thus leveraging the computational resources available on Blue Waters.