Abstract
Radiation plays a unique role in a variety of astrophysical systems because it is both the major observation messenger and a key driver of dynamics. This dissertation discusses physics of astrophysical systems where radiation dominates either the dy- namics or thermodynamics, including radiatively efficient accretion onto black holes and star formation feedback driven by radiation and cosmic rays. The new gener- ation of wide-field variability surveys and multi-band extra-galactic surveys provide new constraints that challenge our theoretical pictures of these sources. Our current understanding is limited by the challenge of modeling radiation forces in complex, evolving geometries. This dissertation is putting effort to use state-of-the-art numer- ical tools, such as Athena++, to improve and expand our models of these systems.
The first three chapters are focusing on star formation feedback. Observations suggest that galactic outflow is ubiquitous, with complex multiphase structures that can be traced by molecular, weakly ionized and neutral gas. With significant mass and momentum loading, galactic outflows may be able to suppress or even quench star formation, acting as an important star formation feedback mechanism. It is still unclear, however, precisely what drives these outflows, although popular ideas include entrainment by hot winds from supernova activity, or non-thermal driving such as radiation or cosmic rays from star forming region.
Chapter 2 investigated radiation pressure’s ability to launch molecular outflow. I study the dynamics of multiphase gas with high temperature contrast with full time-dependent calculation of a two-frequency-band (ultraviolet (UV) and infrared (IR)) radiation field. I find that in contrast to earlier works that focused solely on IR radiation, adding UV component is generally detrimental to cold gas survival during the acceleration, suggesting that radiation pressure acceleration is most promising where IR dominates the spectral energy distribution.
Chapter 3 focuses on cosmic rays (CRs) as another promising candidate for star formation feedback, which are charged particles originate from supernovae activity. I investigate CRs ability to expel ionized outflow from an originally warm, hydrostatic atmosphere. Similar to radiation, I explore the ‘CR Eddington flux’ for galaxies with given surface density and surface star formation rate. The ‘CR Eddington flux’ roughly quantifies CRs ability to disperse gas against gravity and launch outflow. I find that CRs are more likely to be important for galaxies falls on the higher surface density end of Kennicutt-Schmidt relation. Analyzing simulations, I found that the momentum transfer from CR to gas is usually efficient, but the energy transfer de- pends on various factors such as CR flux and the relative importance of diffusion and streaming.
Continuing investigating CR’s role in feedback, I compare multiphase outflows driven by CRs and thermal wind in Chapter 4. Consistent with earlier studies without CRs, I find that when entrained in a hot wind, cold gas can grow instead of evaporate if radiative cooling in turbulent mixing layer is efficient. In contrast, for CR driving, cold gas mass generally decreases even with efficient cooling, albeit at a much slower rate. I show that such different cold gas evolution is related to the intrinsically different nature of CR and thermal pressure gradients near the multiphase interfaces, especially when streaming dominates CR transport. Such different pressure structure can lead to distinct multiphase structure when driven by CR or entrainment in hot wind. For example, they produce outflow with different characteristic column and velocity distributions that can potentially be testing by observations.
In Chapter 5, I discuss radiation’s role in another regime: tidal disruption events, which are transient events powered by accretion near black holes. The disruption of a star by the tidal force from a supermassive blackhole can power a bright transient flare in multiple wavebands that lasts a few weeks to months. However, the origin of detected electromagnetic emission from tidal disruption events is an unresolved puzzle, especially the optical-ultraviolet emission. In this work, I study a poten- tially important pre-peak emission mechanism: stream-stream collision by series of three-dimensional radiation hydrodynamical simulations. I show that for a range of fallback rates, the stream-stream collision can efficiently convert debris’ kinetic en- ergy to radiation, powering prompt emission of ∼ 1042−44erg s−1. n addition, I found that the strong radiation pressure can drive aspherical optical-thick outflow, creating photosphere roughly consistent with pre-peak optical observations. Extending this work, I also introduce the follow-up research project in Chapter 6 to simulate the TDE fallback system in more global calculation domain, aiming to provide insights to the source of early TDE emissions in different wavebands.
“Radiation field and my dear cat Stormy, you both keep me up at night, and I will never fully understand you.”
