Five simulated giant planets migrate inward—but their shared destination is not a universal rule

How gas can move a planet

A protoplanetary disk is the gas-rich material around a young star from which planets can form. An embedded planet pulls on that gas, disturbing its distribution. The gas pulls back, exchanging orbital angular momentum with the planet and potentially moving it inward or outward. The turning effect of those gravitational forces is called torque.

A sufficiently strong disturbance can create a gap: a region with less gas, rather than necessarily an empty ring. The paper distinguishes slower, gap-associated migration from rapid “type III” migration, sometimes called runaway migration. Menezes interprets the simulated behavior as type III because it combines sharp changes in orbital radius with partial gaps.

These processes are relevant to explaining hot Jupiters—giant planets orbiting close to their stars. However, reproducing inward motion in a model is not the same as demonstrating how any particular observed planet arrived there.

What the experiments actually modeled

The study used FARGO3D, a numerical code for simulating gas dynamics and planet–disk interactions. Despite the code’s name and its ability to handle magnetic fields, these calculations were two-dimensional and omitted magnetism. They represented the disk through its surface density, meaning gas mass per unit disk area, and used an isothermal gas prescription rather than modeling heating and cooling in detail.

Each run contained one planet with twice Jupiter’s mass around a star with the Sun’s mass. Starting distances were 0.5, 1.2, 3.2, 5.2 and 11.2 astronomical units, or AU; one AU is approximately the Earth–Sun distance.

The computational grid contained 256 radial zones and 512 angular zones. The disk’s thickness-to-radius ratio was 0.05. Its viscosity, which describes internal transport of momentum, was deliberately set low to encourage rapid migration even for this massive planet.

Crucially, the accretion parameter was zero. Although the title describes a planet “undergoing gas accretion,” the specified planet did not increase its mass. These runs therefore investigate interaction with disk gas, not growth through swallowing it.

Different journeys, similar endpoints

All five simulations developed gaps and ended with the planet inside 0.5 AU. The reported final distances, in order of increasing starting distance, were 0.328, 0.360, 0.361, 0.298 and 0.360 AU. That is a cluster of endpoints, not one identical destination.

The paths also included temporary pauses or plateaus in orbital radius. Apart from the innermost starting case, the runs showed plateaus around 0.8–1.4 AU. The author associates these with changes in partial gaps and the resulting gas interaction. Planets starting farther out showed more pronounced radial variations.

The paper discusses the pauses near Earth’s orbital distance, but this does not establish a connection to habitability or terrestrial-planet formation. Neither was tested by these single-giant-planet calculations.

Why the result needs caution

Five starting positions in one selected model cannot establish a universal stopping distance for giant planets. The study did not systematically vary stellar mass, disk structure, viscosity or planet mass to test whether the endpoint survives those changes. The author explicitly identifies further parameter studies as necessary.

The comparison also has scaling complications. Although starting radius is described as the only varied parameter, the stated conversion from dimensionless surface density to physical units depends on that radius. The time unit changes with starting radius too, so equal output numbers do not represent equal elapsed years across runs.

Timing claims deserve particular care: the conclusion describes migration over years to hundreds of years, while some final snapshots are labeled thousands to tens of thousands of years. Migration episodes need not occupy an entire run, but these statements do not support one clean, universal migration timescale.

Publication context

The supplied arXiv version was posted on November 9, 2026. It identifies the work as already published in Revista Mackenzie de Engenharia e Computação in 2018 and records acceptance in June 2018. It should therefore be understood as an arXiv posting of earlier journal work, not automatically as a newly conducted, unreviewed 2026 study.

The supported takeaway is narrow: under the chosen simplified, low-viscosity conditions, five fixed-mass giant-planet simulations migrated inward and finished at similar distances. Whether that convergence reflects a broadly applicable physical stopping mechanism remains unresolved.