Dynamic Inward/Outward Migrations Across Domains

• Dynamic inward/outward migrations are time-dependent, reversible shifts driven by gravitational forces, angular momentum exchange, and algorithmic processes.
• They underpin structural organization in astrophysical disks, urban migration networks, and computational resource management by creating gaps, pile-ups, and core–periphery arrangements.
• High-resolution simulations and analytical models quantify migration thresholds and reversals, elucidating the diversity in planetary architectures and demographic dynamics.

Dynamic inward and outward migrations refer to the time-dependent, often reversible, changes in the radial locations of bodies—such as planets, planetary embryos, planetesimals, stars, or computational objects—driven by gravitational interactions, angular momentum exchange, dissipative processes, or algorithmic recourse. These migrations are a fundamental organizing process in astrophysical disk systems (planet and star formation), global population movement, and even computer systems, displaying oscillatory, episodic, or stochastic behavior rather than a static one-way drift. Whether occurring in planetary disks, binary star environments, spatial migration networks, or memory hierarchies, dynamic inward/outward migrations often underpin the emergence of characteristic structures such as ‘gaps’, pile-ups, core–periphery organization, or alternating hot/cold resources.

1. Planetesimal-Driven Migration and Oscillatory Core Dispersal

High-resolution N-body simulations of planetesimal-driven migration (PDM) demonstrate that planetary embryos do not remain fixed in their formation zones but instead experience dynamic cycles of inward and outward migration, particularly during the runaway and oligarchic growth phases (Jinno et al., 28 Jan 2026). The radial drift rate of a protoplanet of mass $M_p$ embedded in a sea of planetesimals with surface density $\Sigma_{\rm dust}$ is

$$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$

with PDM torque $\Gamma_{\rm PDM}$ set by the spectrum of planetesimal encounters. The characteristic PDM timescale is

$$\tau_{\rm PDM} = \frac{a_p}{|\dot a_p|} \simeq \frac{M_*\,T_p}{4\pi\,\Sigma_{\rm dust}\,a_p^2}.$$

These simulations reveal that as embryos grow, outward PDM dominates when planetesimal scattering supplies a positive net torque; local depletion of planetesimals eventually inverts the surface density gradient, flipping the torque and inducing inward migration. This results in looping and oscillatory $a_p(t)$ tracks on $10^5$–$10^6$ yr timescales. Mutual repulsion (of order 5 Hill radii) further arranges protoplanet populations into two groups—inner (drifting inward) and outer (moving outward) with a density gap between, thus effectively ‘sorting’ forming planetary cores and enabling delivery of ice-giant–mass embryos ($\sim$3 $M_\oplus$) into the giant planet zone as well as terrestrial-mass cores into the inner disk.

Dynamic Behavior	Physical Process	Typical Timescale
Outward migration	Scattering exterior planetesimals	$\Sigma_{\rm dust}$0–$\Sigma_{\rm dust}$1 yr
Inward migration	Surface density inversion, depletion	$\Sigma_{\rm dust}$2–$\Sigma_{\rm dust}$3 yr
Radial bifurcation	Orbital repulsion $\Sigma_{\rm dust}$4	Few Myr (multi-embryo)

The reversibility and stochasticity of these migrations directly account for radial diffusion, mixing, and the observed diversity in exoplanet system architectures.

2. Disk-Driven Migration in Protoplanetary Environments

Dynamic migration is ubiquitous in gaseous disks, encompassing the entire spectrum from low-mass core migration (Type I), planetesimal-driven regimes, to planet-dominated (Type II) migration of massive planets. Important mechanisms include:

A. Type I and Corotation Torque Reversals

Low-mass planets in radiative disks experience a balance between the negative Lindblad torque and the potentially positive entropy-related corotation torque, with the latter scaling as

$\Sigma_{\rm dust}$5

where $\Sigma_{\rm dust}$6 is set by the entropy gradient. This establishes ‘convergence zones’ towards which isolated planets drift. In systems of multiple embryos, however, resonant trapping increases eccentricity, attenuating the corotation torque and shifting the equilibrium inward, resulting in dynamic stochastic migration—oscillatory, stepwise inward/outward displacement around zero-torque regions (Cossou et al., 2013).

B. Mass- and Gap-Dependent Type II Outward Migration

Recent FARGO3D and multi-dimensional simulations clarify that for sufficiently high gap-depth parameter $\Sigma_{\rm dust}$7, the reversal from inward to outward Type II migration occurs. Typical critical thresholds are $\Sigma_{\rm dust}$8–$\Sigma_{\rm dust}$9 (Scardoni et al., 2022, Scardoni et al., 21 Oct 2025), setting a ‘stalling radius’ $$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$0 toward which planets converge from both sides: $$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$1 Here, the surface density ratio between the 1:2 and 1:3 outer Lindblad resonances, $$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$2, is a key diagnostic for torque sign; outward migration is sustained when this ratio drops below a threshold, typically 0.5–0.8 (Scardoni et al., 21 Oct 2025). For actively accreting planets, azimuthal asymmetries in the circumplanetary flow further enhance positive torques, yielding a migration window in gap-depth parameter space ($$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$3) where outward migration persists (Ida et al., 29 Nov 2025, Pan et al., 26 Nov 2025). Exceeding this window by mass growth reverts the planet to inward drift.

C. Migratory Phase Sequences

Simulations reveal a characteristic sequence: initial inward Type II drift, gap formation and torque reversal, outward migration, and final stalling near $$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$4. This sequence regulates the observed pile-up of giant planets at several astronomical units (Scardoni et al., 21 Oct 2025, Ida et al., 29 Nov 2025).

3. Planetesimal and Population-Scale Migration Exchanges

In multi-planet and planetesimal disk systems, the competition between inward and outward exchanges is essential for explaining radial mixing, resonance capture, and structure formation.

• Neptune and Kuiper Belt objects: Outward migration of Neptune driven by net torque from planetesimal scattering is matched by concurrent inward migration of proto-KBOs via aerodynamic drag (Yeh et al., 2013). The rate equations encapsulate this dual flux:

$$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$5

• Multi-planet chains: Inward–outward exchange, resonant capture, and stochastic migration drive the system toward dynamical equilibria, resonant chains, and clustering (Raymond et al., 2014, Cossou et al., 2013).

4. Population and Large-Scale Migration: Demography and Networks

Dynamic inward/outward migrations are also central to spatial models in human demography, where they manifest as time-dependent net flows between spatial units.

• Dynamic spatial interaction models (SIMs) encode inward (immigration) and outward (emigration) flows at each node via:

$$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$6

where $$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$7 is the instantaneous flow from $$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$8 to $$\dot a_p \simeq \frac{2\,\Gamma_{\rm PDM}}{M_p\,n\,a_p}$$9 and $\Gamma_{\rm PDM}$0 is a transfer rate (Wilkinson et al., 2019). The Two-Trip SIM captures the influence of local neighborhood effects, enabling realistic simulation of multi-center urban clustering via dynamical inward and outward flows.

• Population turnover and diaspora: Flows into and out of a subpopulation (diaspora) follow Poisson processes with arrival and exit rates $\Gamma_{\rm PDM}$1, $\Gamma_{\rm PDM}$2, respectively (Källner et al., 24 Nov 2025). Net change is then $\Gamma_{\rm PDM}$3. This framework quantifies dynamic rejuvenation (younger in-migrants than residents), turnover rates, and geographic patterns such as urban-centric inward migration of young adults and rural net inward flow for seniors.

Migration Type	Mathematical Formulation	Domain
PDM (planets)	$\Gamma_{\rm PDM}$4	Planetary
Diaspora flows	$\Gamma_{\rm PDM}$5	Demographic
SIMs (cities)	$\Gamma_{\rm PDM}$6 per spatial flows $\Gamma_{\rm PDM}$7	Urban systems

• Migration networks: On the global scale, dynamic directed migration networks are encoded in time-dependent weighted adjacency matrices $\Gamma_{\rm PDM}$8. Measures such as in/out-degree, strength, and net migration $\Gamma_{\rm PDM}$9 track dynamic inward/outward roles (sources/sinks), with hyperbolic embedding revealing persistence and reorganization of migration 'cores' and peripheries over time (Gou et al., 2020).

5. Computational Systems: Memory and Resource Management via Dynamic Migrations

In computer systems, dynamic inward/outward migrations appear in DRAM power management, bin-packing algorithms, and virtual machine placement.

• DRAM page migration: Hot (frequently accessed) pages are migrated 'inward' to hot ranks (continuous access) while cold pages move 'outward' to ranks intended for idling (Lu et al., 2014). Formal grouping is based on recency/frequency metrics, with migrations planned via bipartite matching and scheduled through Eulerian cycle decomposition. This maximizes power-state residency efficiency, reducing energy-delay squared by 60–65%. Quantitatively, disabling dynamic migrations alone degrades $$\tau_{\rm PDM} = \frac{a_p}{|\dot a_p|} \simeq \frac{M_*\,T_p}{4\pi\,\Sigma_{\rm dust}\,a_p^2}.$$0 by 17–22%.
• Dynamic bin packing: Dynamic inward/outward migrations (item transfers) are leveraged to optimize total active bin-time. There is a sharp threshold: sublinear ($$\tau_{\rm PDM} = \frac{a_p}{|\dot a_p|} \simeq \frac{M_*\,T_p}{4\pi\,\Sigma_{\rm dust}\,a_p^2}.$$1) migrations yield no improvement relative to zero migrations (competitive ratio grows as $$\tau_{\rm PDM} = \frac{a_p}{|\dot a_p|} \simeq \frac{M_*\,T_p}{4\pi\,\Sigma_{\rm dust}\,a_p^2}.$$2), but linear ($$\tau_{\rm PDM} = \frac{a_p}{|\dot a_p|} \simeq \frac{M_*\,T_p}{4\pi\,\Sigma_{\rm dust}\,a_p^2}.$$3) migrations attain $$\tau_{\rm PDM} = \frac{a_p}{|\dot a_p|} \simeq \frac{M_*\,T_p}{4\pi\,\Sigma_{\rm dust}\,a_p^2}.$$4-approximation, with an explicit trade-off (Mellou et al., 2024). Migration algorithms trigger inward transfer only when an outward (source) bin's load falls below a threshold, preventing oscillatory 'ping-pong' exchange.

6. Population III Star/Binary and Stellar Cluster Migrations

Outward migration phenomena are not limited to planetary contexts. In Population III binary star formation, binaries can dynamically migrate outward via angular momentum transfer from rapidly accreted high-j gas onto circumstellar minidiscs (Park et al., 2023). The net torque and migration rate are determined by both gravitational torques from the circumbinary disk and the anisotropy of mass accretion. Equal-mass binaries with high accretion rates ($$\tau_{\rm PDM} = \frac{a_p}{|\dot a_p|} \simeq \frac{M_*\,T_p}{4\pi\,\Sigma_{\rm dust}\,a_p^2}.$$5) favor outward separation growth. Conversely, lower accretion or unequal masses, and strong radiative feedback, drive rapid inward migration and fragment merging.

7. Synthesis and Theoretical Implications

Dynamic inward/outward migrations are generic phenomena arising from local or global imbalances of angular momentum, mass, or resource flows. Whether induced by scattering, dissipative drag, torque reversals, migration traps, resonance crossing, or algorithmic reallocation, these migrations are responsible for self-organization, mixing, gap formation, and demographic rejuvenation in disparate physical and abstract systems. Multiple studies highlight the reversibility and episodic character of these processes, the presence of migration windows or stalling radii, the importance of migration thresholds (e.g., critical $$\tau_{\rm PDM} = \frac{a_p}{|\dot a_p|} \simeq \frac{M_*\,T_p}{4\pi\,\Sigma_{\rm dust}\,a_p^2}.$$6, $$\tau_{\rm PDM} = \frac{a_p}{|\dot a_p|} \simeq \frac{M_*\,T_p}{4\pi\,\Sigma_{\rm dust}\,a_p^2}.$$7, or demographic turnover rates), and the central role of feedback between inflow and outflow in shaping system-scale structure (Jinno et al., 28 Jan 2026, Scardoni et al., 21 Oct 2025, Ida et al., 29 Nov 2025, Källner et al., 24 Nov 2025).

In population dynamics, combined modeling of inward and outward flows exposes previously hidden replacement dynamics and turnover. In astrophysics, such dynamic migrations are essential for explaining the present distribution of planetary and satellite systems as well as the demography of debris disks. In computer science, migration scheduling under dynamic workloads offers provable resource-utilization efficiencies and underlines the critical value of judicious, rather than arbitrary, recourse.

Ultimately, dynamic migration—via the continual interplay of inward and outward phases—serves as an organizing principle across astrophysical, demographic, and computational domains.