No Arabic abstract
The formation of the first planetesimals and the final growth of planetary cores relies on the abundance of small pebbles. The efficiencies of both the streaming instability (SI) process, suggested to catalyze the early growth of planetesimals, and the pebble-accretion process, suggested to accelerate the growth of planetary cores, depend on the sizes of solids residing in the disk. In particular, these processes were found to be sensitive to size distribution of solids, and efficient planetesimal formation and growth through these channels require a limited pebble size distribution. Here we show that aeolian erosion, a process that efficiently grinds down boulders into a mono-sized distribution of pebbles, provides a natural upper limit for the maximal pebble sizes (in terms of their Stokes number). We find the dependence of this upper limit on the radial separation, disk age, turbulence strength, and the grain-size composition of the boulders in the disk. SI is favorable in areas with a Stokes number less than 0.1, which is found in the inner sub-astronomical-unit regions of the disk. This upper limit shapes the size distribution of small pebbles and thereby catalyzes the early onset of planetesimal formation due to SI, and the later core accretion growth through pebble accretion.
We present local simulations that verify the linear streaming instability that arises from aerodynamic coupling between solids and gas in protoplanetary disks. This robust instability creates enhancements in the particle density in order to tap the free energy of the relative drift between solids and gas, generated by the radial pressure gradient of the disk. We confirm the analytic growth rates found by Youdin & Goodman (2005) using grid hydrodynamics to simulate the gas and, alternatively, particle and grid representations of the solids. Since the analytic derivation approximates particles as a fluid, this work corroborates the streaming instability when solids are treated as particles. The idealized physical conditions -- axisymmetry, uniform particle size, and the neglect of vertical stratification and collisions -- provide a rigorous, well-defined test of any numerical algorithm for coupled particle-gas dynamics in protoplanetary disks. We describe a numerical particle-mesh implementation of the drag force, which is crucial for resolving the coupled oscillations. Finally we comment on the balance of energy and angular momentum in two-component disks with frictional coupling. A companion paper details the non-linear evolution of the streaming instability into saturated turbulence with dense particle clumps.
Based on the interstellar CO/H2 ratio, carbon monoxide-based censuses of protoplanetary disks in Lupus, sigma Orionis, and Chamaeleon I found no disks more massive than the minimum-mass solar nebula, which is inconsistent with the existence of exoplanets more massive than Jupiter. Observations and models are converging on the idea that ionization-driven chemistry depletes carbon monoxide in T-Tauri disks. Yet the extent of CO depletion depends on the incident flux of ionizing radiation, and some T-Tauri stars may have winds strong enough to shield their disks from cosmic rays. There is also a range of X-ray luminosities possible for a given stellar mass. Here we use a suite of chemical models, each with a different incident X-ray or cosmic-ray flux, to assess whether CO depletion is a typical outcome for T-Tauri disks. We find that CO dissociation in the outer disk is a robust result for realistic ionization rates, with abundance reductions between 70% and 99.99% over 2 Myr of evolution. Furthermore, after the initial dissociation epoch, the inner disk shows some recovery of the CO abundance from CO2 dissociation. In highly ionized disks, CO recovery in the inner disk combined with depletion in the outer disk creates a centrally peaked CO abundance distribution. The emitting area in rare CO isotopologues may be an indirect ionization indicator: in a cluster of disks with similar ages, those with the most compact CO isotopologue emission see the highest ionization rates.
The streaming instability is a leading candidate mechanism to explain the formation of planetesimals. Yet, the role of this instability in the driving of turbulence in protoplanetary disks, given its fundamental nature as a linear hydrodynamical instability, has so far not been investigated in detail. We study the turbulence that is induced by the streaming instability as well as its interaction with the vertical shear instability. For this purpose, we employ the FLASH Code to conduct two-dimensional axisymmetric global disk simulations spanning radii from $1$ au to $100$ au, including the mutual drag between gas and dust as well as the radial and vertical stellar gravity. If the streaming instability and the vertical shear instability start their growth at the same time, we find the turbulence in the dust mid-plane layer to be primarily driven by the streaming instability. It gives rise to vertical gas motions with a Mach number of up to ${sim}10^{-2}$. The dust scale height is set in a self-regulatory manner to about $1%$ of the gas scale height. In contrast, if the vertical shear instability is allowed to saturate before the dust is introduced into our simulations, then it continues to be the main source of the turbulence in the dust layer. The vertical shear instability induces turbulence with a Mach number of ${sim}10^{-1}$ and thus impedes dust sedimentation. Nonetheless, we find the vertical shear instability and the streaming instability in combination to lead to radial dust concentration in long-lived accumulations which are significantly denser than those formed by the streaming instability alone. Thus, the vertical shear instability may promote planetesimal formation by creating weak overdensities that act as seeds for the streaming instability.
We present simulations of the non-linear evolution of streaming instabilities in protoplanetary disks. The two components of the disk, gas treated with grid hydrodynamics and solids treated as superparticles, are mutually coupled by drag forces. We find that the initially laminar equilibrium flow spontaneously develops into turbulence in our unstratified local model. Marginally coupled solids (that couple to the gas on a Keplerian time-scale) trigger an upward cascade to large particle clumps with peak overdensities above 100. The clumps evolve dynamically by losing material downstream to the radial drift flow while receiving recycled material from upstream. Smaller, more tightly coupled solids produce weaker turbulence with more transient overdensities on smaller length scales. The net inward radial drift is decreased for marginally coupled particles, whereas the tightly coupled particles migrate faster in the saturated turbulent state. The turbulent diffusion of solid particles, measured by their random walk, depends strongly on their stopping time and on the solids-to-gas ratio of the background state, but diffusion is generally modest, particularly for tightly coupled solids. Angular momentum transport is too weak and of the wrong sign to influence stellar accretion. Self-gravity and collisions will be needed to determine the relevance of particle overdensities for planetesimal formation.
Damping of the previously discovered resonant drag instability (RDI) of dust streaming in protoplanetary disc is studied using the local approach to dynamics of gas-dust perturbations in the limit of the small dust fraction. Turbulence in a disc is represented by the effective viscosity and diffusivity in equations of motion for gas and dust, respectively. In the standard case of the Schmidt number (ratio of the effective viscosity to diffusivity) Sc = 1, the reduced description of RDI in terms of the inertial wave (IW) and the streaming dust wave (SDW) falling in resonance with each other reveals that damping solution differs from the inviscid solution simply by adding the characteristic damping frequency to its growth rate. RDI is fully suppressed at the threshold viscosity, which is estimated analytically, first, for radial drift, next, for vertical settling of dust, and at last, in the case of settling combined with radial drift of the dust. In the last case, RDI survives up to the highest threshold viscosity, with a greater excess for smaller solids. Once Sc eq 1, a new instability specific for dissipative perturbations on the dust settling background emerges. This instability of the quasi-resonant nature is referred to as settling viscous instability (SVI). The mode akin to SDW (IW) becomes growing in a region of long waves provided that Sc > 1 (Sc < 1). SVI leads to an additional increase of the threshold viscosity.