Date of Award
Doctor of Philosophy (PhD)
This thesis presents a theoretical investigation of supermassive black holes and the quasars they power through cosmological hydrodynamic simulations on petascale supercomputers. As the size of simulations increase, visualization and interaction with the data become difficult. We developed interactive visualization software on top of existing image deep-zoom and tagging infrastructure libraries and platforms, specifically in the context of cosmological hydrodynamic simulations. The visualization tools were instrumental in establishing the cold flow fueling model of high redshift quasar growth. We then proceed to further study this growth model with high resolution zoom-in re-simulations, and report that the cold flows feeding supermassive blackholes are numerically stable. We studied the morphology of HII regions with a parallel implementation of the ray tracing scheme SPHRAY, and found that the HII regions due to high redshift quasars are smoother and more extended that those due to galaxies. Finally, we developed a set of tools to produce and verify mocked correlated quasars and Lyman-α forest. We fit the linear theory correlation function of the Quasar-Quasar and Forest-Forest auto-correlation and Quasar-Forest cross-correlation of the mocked transmission fraction and quasar locations up to 160 Mpc/h.
Feng, Yu, "Petascale Cosmological Hydrodynamic Simulation of Quasars" (2014). Dissertations. 395.