Encounters Between M33 and Present-Day M31 Satellites Hint at a Previous Group Accretion

This work investigates whether two known Andromeda (M31) satellites, Pisces (LGS 3) and Andromeda XVI, have interacted with M33, M31’s most massive satellite. $Λ$CDM predictions imply a handful of satellite galaxies around M33, yet few M33 satellites have been found and confirmed despite its high mass. We use proper motions combined with backward orbit integration in a semi-analytic potential to constrain plausible interaction scenarios for Pisces and And XVI. Both dwarfs are currently M31 satellites, defined as being inside its virial radius. However, our results show that, in our fiducial mass models, 42% (And XVI) and 60% (Pisces) of dwarf orbits support that they were previously satellites of M33 (i.e., once inside its virial radius). Both dwarfs had fly-by encounters with M33 at relative velocities greater than M33’s escape speed within the past 1-2 Gyr. In over 70% of orbits, Pisces and And XVI also had a close approach with each other post-M33 interaction and share an orbital plane, suggesting possible past group accretion. We explore a range of mass combinations for M31 and M33, finding that this primarily regulates the likelihood that the dwarfs were satellites of M33 in the past, while upholding conclusions of recent flybys about M33. These close interactions provide new evidence for past satellite exchange and/or group infall scenarios between M31 and M33. Such interactions also affect comparisons to observational surveys that define satellites primarily by their distance relative to host galaxies.

💡 Research Summary

This paper investigates whether the two well‑studied dwarf galaxies Pisces (LGS 3) and Andromeda XVI, currently classified as satellites of the Andromeda galaxy (M31), were once satellites of M33, the most massive satellite of M31. ΛCDM cosmology predicts that a galaxy of M33’s mass should host a handful of its own dwarf companions, yet observational searches have identified only a few candidate M33 satellites, leaving a tension between theory and data.

To address this, the authors combine recent proper‑motion measurements from multi‑epoch Hubble Space Telescope imaging (three epochs for Pisces, two for And XVI) with line‑of‑sight velocities and distance estimates, constructing full six‑dimensional phase‑space vectors for each dwarf. They embed these dwarfs in a semi‑analytic, time‑dependent gravitational potential that includes the Milky Way, M31, M33, the Large Magellanic Cloud, and the dwarfs themselves. The massive hosts are modeled with three‑component (halo, disk, bulge) potentials, with dynamical friction and adiabatic halo contraction accounted for via the CONTRA code. Adopted virial masses are 2 × 10¹² M⊙ for M31, 1 × 10¹² M⊙ for the Milky Way, 1 × 10¹¹ M⊙ for M33, and 2.5 × 10¹¹ M⊙ for the LMC.

Using a Leapfrog integrator, the authors integrate the equations of motion backward for 6 Gyr, generating 1,000 Monte‑Carlo realizations for each dwarf to propagate observational uncertainties. They then examine whether the dwarfs ever entered the virial radius of M33 (Rvir ≈ 161 kpc) and, if so, whether they were bound (velocity below the local escape speed).

Key findings:

1. Current status – Both Pisces and And XVI are presently inside M31’s virial radius (Rvir ≈ 329 kpc) and are on their first infall into M31’s halo. Only a small fraction of Monte‑Carlo orbits show a prior pericenter with M31 (≈2 % for Pisces, ≈28 % for And XVI).

2. Past M33 interaction – In the fiducial mass model, 60 % of Pisces orbits and 42 % of And XVI orbits entered M33’s virial sphere within the last 1–2 Gyr. However, the relative velocities at those passages exceed M33’s escape speed, indicating that the dwarfs were not long‑lived bound satellites but rather experienced high‑speed fly‑bys.

3. Joint dwarf–dwarf encounter – More than 70 % of the realizations show that after their respective M33 fly‑bys, Pisces and And XVI approached each other closely, sharing a common orbital plane (primarily confined to the M31‑centric x‑z plane). This alignment is reminiscent of the coherent orbital poles seen in group infall scenarios, such as the satellites of the Large Magellanic Cloud.

4. Mass dependence – Varying the assumed masses of M31 and M33 strongly influences the probability of a past M33‑satellite status. Higher M33 mass or lower M31 mass both increase the fraction of orbits that satisfy the satellite criterion, underscoring the sensitivity of such dynamical histories to host‑halo mass uncertainties.

5. Implications – The results provide concrete dynamical evidence that M33 likely possessed a more substantial satellite system in the past, part of which may have been stripped during its recent close passage with M31 (the “fly‑by” scenario). The apparent exchange of satellites between M33 and M31, together with the possible group infall of Pisces and And XVI, suggests that satellite‑to‑satellite transfer and group accretion are viable channels for building the present‑day dwarf population in the Local Group. Moreover, the study highlights that defining satellites solely by distance relative to a host can be misleading; a dynamical definition incorporating virial radii and escape velocities is more appropriate.

In summary, by leveraging precise proper motions and a realistic multi‑body potential, the authors demonstrate that Pisces and Andromeda XVI likely experienced recent, high‑velocity encounters with M33 and subsequently with each other, supporting a scenario of past group accretion or satellite exchange. This work bridges the gap between ΛCDM predictions of M33’s satellite abundance and the paucity of observed companions, and it calls for deeper, kinematically complete surveys of the M33 halo to fully resolve its dwarf satellite census.