Analysing and predicting chaos from shocks and radiative cooling in simulations of cosmic structure formation

Abstract

It has long been appreciated that galaxy formation involves physics that can be subject to chaos. Numerical simulations introduce an additional source of noise in the form of approximations and floating point arithmetic errors. To date, there have been preliminary investigations into quantifying and characterizing the effects of both physical and numerical factors on simulation results, however, focusing on individual physical processes, such as shocks or radiative cooling, has been somewhat overlooked. In this thesis, we provide an analysis of the impact of shocks and radiative cooling in simulations of cosmic structure formation, by investigating a combination of phase space separation and density-phase space correlation. We have turned off other processes, such as star formation and feedback, to be able to focus on hydrodynamics with cooling alone. We find that regardless of cooling, galaxy mergers exhibit chaotic properties, however, cooling enables faster and stronger response to small initial changes. An initial hypothesis that high density regions should be more chaotic than low density regions is also shown to breakdown, especially at later times during the merger. Overall the development of differences in solutions is complex and involves distinct physical processes as well as mass scales.