Planet formation and the early evolution of self-gravitating protoplanetary discs

Abstract

When a Giant Molecular Cloud (GMC) collapses to form a stellar core, conservation of angular momentum will lead to the formation of a protoplanetary disc, with an initial mass potentially of the order of its stellar host. If a massive disc forms, then the disc’s self-gravity will play a crucial role in the earliest stages of its evolution; driving its viscous evolution, and potentially leading to the formation of wide orbit, giant planets and brown dwarfs through disc fragmentation. I begin this thesis by placing improved constraints on the conditions required for disc fragmentation, specifically focusing on how the disc’s environment may influence its evolution and eventual fate. Recent results from direct imaging surveys suggest that wide orbit giant planets and brown dwarfs are found more frequently around higher mass stars. I use Smoothed Particle Hydrodynamics (SPH) simulations to show that a disc’s susceptibility to fragmentation is dependent on the mass of its host star. I demonstrate that discs around higher mass stars may fragment for lower disc-to-star mass ratios, making them favourable sites for the formation of wide orbit, massive objects, such as those found in direct imaging surveys. Low mass stars may support high mass discs, in principle providing large reservoirs of material for core accretion planet formation. Results from direct imaging surveys also find that stars hosting close in giant planets or brown dwarfs display an excess of outer binary companions, with indications that some of these objects may have formed through the gravitational instability (GI). I use SPH to simulate a suite of self-gravitating discs with a binary companion, and show that there is a narrow region of parameter space where intermediate separation companions may trigger fragmentation. Short separation encounters are destructive, whilst wide orbit companions have little effect. The range of binary separations found to favour the formation of short period, giant planets is consistent with results from direct imaging surveys. Although numerical models suggest that GI may dominate a disc’s early evolution, it is still unclear from observations whether massive, self-gravitating discs exist in nature. Recent high-resolution infrared imaging of protoplanetary discs have given rise to unparalleled observations of their substructure, including rings, gaps and spirals, providing us with crucial insights to the earliest stages of planet formation. Observations of the protoplanetary disc surrounding AB Aurigae have revealed the possible presence of two massive planets in the process of forming. The young measured age for the system places strict time constraints on the planet’s formation histories. I use analytic core accretion models to show that their expected core accretion formation timescales are longer than the system’s current age. Using SPH and viscous evolution models of self-gravitating discs, I show that a proto-AB Aurigae disc could have been massive enough to fragment in the past, with typical fragment masses consistent with the masses of the protoplanets which have been observed in the disc. Finally, I use Monte Carlo radiative transfer models to generate observational predictions of self-gravitating discs using ALMA. I develop an existing 3D semi-analytic model to include a prescription for dust trapping in the disc’s spirals. I make predictions about the disc properties which may drive spirals that could be visible to ALMA, in particular focusing on the impact of dust trapping. I also use these models to analyse 3 discs from the DSHARP survey, and discuss the plausibility of their observed spirals being the result of GI.