Skip to Content

Realistic Simulations of the Intergalactic Medium: The Search for Missing Physics - Part 2

Michael Norman, University of California, San Diego

Usage Details

Michael Norman, James Bordner, Alexei Kritsuk, Britton Smith, Kyungjin Ahn, Azton Wells

The intergalactic medium (IGM) is the hot, X-ray emitting gas that permeates the space between galaxies. Currently, standard numerical simulation cannot match several key observational features of the IGM. The mismatch between the simulations and observation data is much too large to be caused by observational errors, and is seen in multiple comparisons using different statistics, and by different groups. This project proposes to resolve this open question in astrophysics by running very high resolution simulations on the Blue Waters supercomputer to make more accurate and robust estimates of the IGM. Furthermore, the simulations and codes will be publicly available to allow others to make their own estimates from existing or new observations.

The project will simulate the hydrogen Lyman alpha forest (LYAF) using the newly developed extreme scale branch of the Enzo called Enzo-P. Enzo-P implements the rich suite of physics models from the widely used Enzo community code on top of a new, highly scalable AMR infrastructure called Cello. Both Enzo-P and Cello have been developed by the PIs group at UCSD. The combination of Enzo-P and Cello will permit for the first time the use of AMR throughout a large cosmological volume, removing the existing tradeoff between large volume statistics and high local resolution that characterized previous Enzo simulations. The project will apply this new capability to carry out simulations of the LAF with sufficiently high resolution to capture the gaseous halos of galaxies which contribute significantly to intergalactic absorption, while at the same time surveying a statistically significant volume of the universe. Moreover, the project will construct synthetic LYAF spectra and analyze their statistical properties to see whether the inclusion of high column density absorbers in the model improves agreement observations.