What is the contribution of gravitational infall on the mass assembly of star-forming clouds? A case study in a numerical simulation of the interstellar medium

TL;DR

This study quantifies the role of gravitational infall in assembling star-forming clouds using a stratified-box ISM simulation that includes SN and photoionization feedback. By deploying tracer particles and time-averaging accelerations, it separates gravity-driven inflow from turbulence and external potential effects, finding that gravity contributes a modest fraction (~10–30% at 100 pc) of mass inflow and up to ~45% within the densest molecular gas, while the large-scale linewidth is not gravity-dominated. The results show a gradual shift toward gravity-dominated inflow as gas densifies, suggesting cloud assembly is primarily turbulence-driven at large scales but gravity becomes increasingly important at higher densities inside clouds. These findings bridge turbulent-support and gravity-dominated perspectives, with implications for analytic SFR models and the interpretation of molecular cloud linewidths, while highlighting the dominant role of SN-driven turbulence in cloud assembly and the eventual prominence of self-gravity in the densest regimes.

Abstract

Star formation in galaxies is a complex phenomenon occurring on a very wide range of scales, and molecular clouds are at the heart of this process. The formation of these structures and the subsequent collapse of the gas within them to form new stars remain unresolved scientific questions. In particular, the role and importance of gravity at between the disk scale height and prestellar cores (100 to 0.01 pc) are still topics of debate. In this work, we conduct a case study examining the mass assembly and evolution of a giant molecular cloud complex in a numerical stratified-box simulation of the interstellar medium with photo-ionizing and supernova driving and resolving down to scales $\gtrsim 1$ pc and densities up to $10^3$ cm$^{-3}$. By introducing tracer particles to precisely track the forces acting on the gas during its evolution towards and within the clouds, we are able to quantify how much of the mass inflow is driven by the self-gravity of the gas and the gravity from the stellar disk. We find that up to 20% of the gas is gravity-driven at a scale of 100 pc, contributing 10% of the inflow from the warm to the cold phase and 20% from the cold phase to the individual molecular clouds, reaching up to 45% inside the molecular gas, at densities $\gtrsim 400$ cm$^{-3}$. However, at the 100 pc scale, the contribution of gravity-driven gas on the linewidth is negligible. We conclude that the bulk of the gas motions assembling the clouds in our simulation are caused by the supernova-driven supersonic turbulence, which results in locally convergent flows, with a small contribution from the stellar gravitational potential.

Figures (15)

• Figure 1: Evolution of the surface density of the full kpc$^3$ box from t = 64 Myr and t = 94 Myr. We study the timespan between t = 74 Myr and t = 84 Myr, while the main over-dense structure already formed but is not destroyed by feedback yet. White dots represent the position of sink particles.
• Figure 2: Overall face-on view of the simulation domain, centered on the main overdensity. Top: Surface density over a 100 pc vertical thickness at 74 Myr (left) and 84 Myr (right) of the stratified ISM box simulation, corresponding to the first and final snapshots we use for our case study. The map is centered around the minimum of the potential. The streamlines depict the direction of the velocity field. The sink particles (representing stellar clusters) are also depicted, with age coded in color. Bottom: Close-up view around the minimum of the gravitational potential for both snapshots. Here the mean density in a 20-pc thick slice is depicted. The contours lines represent the gravitational potential, in units of 10$^{7}$ cm$^2~$s$^{-2}$, and in a gauge where its minimum is set to zero.
• Figure 3: Average density over a 20-pc thick slice around the minimum of the potential at the beginning (left) and the end (right) of the studied time-span. The dashed region shows where at least 20% of the gas tracers are gravity-driven. The top panels show the face-on view while the bottom panels depict an edge-on view.
• Figure 4: Enclosed mass within 100 pc of the minimum of the potential that is over a given density threshold at the starting time $t_s$ (solid line) and the end $t_e$ (dash-dotted) line. To help the interpretation, the sink formation threshold (vertical orange dotted line) is added, as well as the density thresholds above which the gas is expected to be in the CNM (light blue shaded area) or molecular (dark blue shaded area) phase components, based on drainePhysicsInterstellarIntergalactic2011).
• Figure 5: Fraction of the total mass in tracer particles in gravity-driven particles during the 10-Myr time span, as a function of distance from the minimum of the potential.