A longstanding assumption is that planetary growth commences after the infall of gas and solids to a circumstellar disk ends, with disk infall and planet accretion traditionally modeled as two separate phases. While conceptually and computationally convenient, this division may not always be physically valid: indeed, there is substantial observational evidence that accretion commences early and even during infall in some systems.
We are developing a first generation of simulations to assess the effects of planet accretion during late disk infall on resulting system properties. In Rufu & Canup (2025), we focused on compact systems, whose short orbital timescales imply that once > km-sized planetesimals form, full planet accretion proceeds rapidly on timescales shorter than or comparable to infall timescales. We simulated planet growth within a disk supplied by an infall with small centrifugal radius, rc ~ 0.1 to 0.5 au. In disk regions undergoing infall (i.e., for r < rc,), planet masses are set by a balance between accretion of infalling solids and Type I migration. This produces a similar planet size within each system, consistent with the observed “peas-in-a-pod’’ structure, which can also be explained by prior, post-infall formation models.

Figure 1: Estimated total mass of transiting compact systems, Mtot, scaled to the stellar mass, M*, for compact systems having ≥3 known planets that orbit a single star within a<0.5 au (circle markers, blue box). Points are ordered left-to-right by ascending stellar mass. For cases without mass estimates, we use the observed planet radius, increase the estimated radius uncertainty by a factor of 2, and then apply a radius vs. mass relation. Light [medium] blue circles are systems with all [some] planetary masses estimated from this relation, while dark blue circles are systems with measured planetary masses. Over a wide range of stellar masses, compact multi-planet systems display a common mass ratio, with 90% of systems having 3 x10-5 < (Mtot/M*) < 3 x 10-4. This mass ratio is more similar to that of the gas giant satellite systems (square markers, yellow box) than to the inner or outer planets in our Solar System (triangle markers, red box).
More notably, Rufu & Canup (2025) find that accretion during infall explains two traits of compact systems that are not easily explained by standard, post-infall models. First, compact systems display a remarkably consistent ratio between the total planetary mass and the stellar mass, with this ratio being few times 10-5 to 10-4 across systems whose stellar masses vary by an order-of-magnitude (Figure 1). Why such a preferred mass ratio would exist for varied stellar masses and disk evolutions has been a mystery. We show that accretion during infall regulates a compact system to have this common mass ratio for a wide range of disk and infall conditions (Figure 2). Second, the mass of compact system planets shows an unusually weak dependence on stellar metallicity, in contrast to, e.g., gas giants. For a standard post-infall model, planet masses would generally be proportional to metallicity, in contrast to the observational trend. We find that accretion during infall yields planet and planet system masses that have only a weak dependence on metallicity, providing an explanation to this long-standing problem.
We will discuss observational implications of planet accretion during infall and key areas for further advancement. The latter include a better understanding of where and when early planetesimals may form, particularly during the late stages of infall.

Figure 2: Results of compact planet system accretion simulations with varied disk and infall properties. Final planetary system mass scaled to the stellar mass as a function of (ae/f) (a is the viscosity parameter, e is the fraction of infalling solids incorporated into planets, and f is the infall gas-to-solids ratio). The infall rate decays with timescale tin = 5 x 105 yr, while the gas disk disperses over a longer timescale, tg= 1.3 to 2tin (colors, legend). The simulations assume either an inner disk cavity (triangles) or no cavity (circles). Grey region shows range for 90% of observed compact systems shown in Figure 1. Dashed lines show analytical predictions for the no-cavity case. Horizontal bars show plausible viscosity ranges, assuming (f/e) = 100.
Rufu, R. and R. M. Canup (2025) “Origin of compact exoplanetary systems during disk infall” Nature Communications, 16, 4853.