Modeling the recent interactions between the Magellanic Clouds and Milky Way

Bethany Garver; David Nidever; Nathan Deg; Victor Debattista

REVIEW 4 major objections 5 minor 1 cited by

Modeling the recent interactions between the Magellanic Clouds and Milky Way

T0 review · 4 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A best-matching N-body model of the Magellanic Clouds over the past 2.5 Gyr finds two close Cloud–Cloud encounters and attributes the LMC's U-shaped warp to the recent one, 140 Myr ago.

desk verdict A credible, honest N-body reconstruction of recent LMC–SMC–MW interactions with a plausible 140 Myr encounter, but the headline claim that this encounter causes the observed U-shaped warp is not robust to plausible LMC halo concentrations—and the feature matches are post-hoc, not fitted. read the letter →

arxiv 2602.05021 v1 pith:F3OF6MXB submitted 2026-02-04 astro-ph.GA

Bethany Garver , David Nidever , Victor Debattista , Nathan Deg This is my paper

classification astro-ph.GA

keywords MagellanicCloudsLMCdiscwarpgalaxyinteractionsN-bodysimulationsgeneticalgorithmbendingwavesSMCtidalexpansionstellarsubstructure

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper is trying to establish that a single recent collision between the Large and Small Magellanic Clouds—about 140 million years ago—shaped several of the most prominent observed features of the system, most notably the U-shaped warp of the LMC's disc. It also reconstructs an earlier close passage roughly 940 Myr ago. The authors run about 8,000 N-body simulations with a genetic algorithm to find initial conditions that bring the simulated Clouds to their observed positions and velocities today; their best model simultaneously produces the LMC's ring-shaped overdensity, the LMC's northern stream, the SMC's tidal expansion, and the eastern SMC's two distance components. A sympathetic reader would care because it offers a concrete, testable timeline for the closest interacting galaxy pair we can study in detail, and it makes a prediction: the outer LMC disc should be vertically oscillating with a period of about 600 Myr.

What carries the argument

The load-bearing machinery is a genetic algorithm coupled to live N-body simulations (each with stellar and dark matter particles) that searches the initial positions and velocities of the LMC and SMC 2.5 Gyr ago. Unlike analytical orbit integration, the N-body treatment lets the galaxies deform, so tidal stripping and dynamical friction are captured self-consistently during close passages. The warp mechanism itself is a vertical bending wave: the recent direct impact displaces the outer disc, which then oscillates with a period of roughly 600 Myr, and the observed U-shape is the inner part of that wave.

What would settle it

Measure the three-dimensional kinematics of stars in the outer LMC disc (beyond ~7–8 kpc). If the outer disc is not oscillating vertically with a period of roughly 600 Myr and an amplitude of ~1 kpc, or if the warp is found to pre-date 140 Myr (e.g., in the age distribution of warp stars), then the claim that the recent SMC collision caused the warp is falsified. Alternatively, an independent measurement of the LMC's dark matter halo scale radius showing it to be as compact as model F (~13 kpc) would predict only small ripples, contradicting the observed U-shaped warp.

Watch

Extended reading notes

Core claim

The central discovery is that the LMC's current U-shaped warp is a bending wave excited by a direct collision with the SMC about 140 Myr ago, not a static tidal distortion. In the best-fitting simulation, the LMC and SMC have two close passages (940 Myr and 140 Myr ago); the second sends vertical ripples outward through the LMC disc with mean amplitude ~1.3 kpc, and the observed southern and northern warps correspond to the inner edge of these oscillations. The same simulation reproduces the LMC ring overdensity as a temporary overlap of two spiral arms, the SMC's radial tidal expansion with slope ~10 km/s/kpc, and the eastern SMC distance bimodality as a close retrograde tidal arm stripped

Load-bearing premise

The LMC's dark matter halo must be neither too compact nor too diffuse in just the right way: with a compact halo the recent SMC encounter produces only small ripples, and with a diffuse halo it produces an S-shaped warp, so the claim that the 140 Myr collision caused the observed U-shaped warp hinges on the real LMC halo falling in between.

Editorial extensions

If this is right

The interaction times are pinned to ~940 Myr and ~140 Myr before the present, giving a timeline that other observations (e.g., star formation histories, stellar populations) can be checked against.
The LMC's outer disc should exhibit vertical oscillations with a ~600 Myr period and amplitude ~1.3 kpc; deep imaging or proper motions of outer-disc stars can test this directly.
The ring-shaped overdensity is a transient feature formed by two overlapping spiral arms; it should be expanding outward and will dissipate, meaning the LMC's appearance changes on a few hundred Myr timescale.
The SMC's radial expansion and the eastern distance bimodality both derive from the same recent collision, so measurements of SMC kinematics and composition across the eastern side probe the collision geometry.
The absence of the southern hooks and the SMC Northern Overdensity in the best model suggests those structures formed more than 2.5 Gyr ago, motivating earlier-time simulations.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the warp is a bending wave, the LMC's vertical velocity field should show a coherent radial pattern (rising on one side, falling on the other) that can be looked for in Gaia proper motions and future spectroscopic surveys.
The two-interaction timeline predicts that the LMC's star formation may have been triggered twice in the past ~1 Gyr; comparing age distributions of LMC and SMC stellar populations could corroborate or reject the 940 Myr encounter.
The halo-concentration sensitivity of the warp suggests that measuring the LMC's dark matter halo shape (e.g., through kinematics of the outer disc or stellar streams) is a direct way to validate the model; if the halo is compact, the proposed origin of the U-shaped warp is in trouble.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Worth your time if you care about Magellanic Cloud dynamics. The genuinely new thing here is a full three-galaxy, live N-body search over ~8000 genetic-algorithm runs, plus a public data release. That is a real step beyond the old test-particle GA work (Ruzicka et al. 2010) and beyond fixed-potential suites. The best model lands on two close LMC–SMC encounters, at ~940 and ~140 Myr ago, and it reproduces several observed substructures in a single simulation: the LMC ring, the northern stream, the SMC's tidal expansion, and the eastern distance bimodality. I found the paper refreshingly honest: it reports its failures (no southern hooks, no SMCNOD), gives the residuals, and makes a concrete, falsifiable prediction about vertical oscillations in the outer LMC.

The soft spots are real but not fatal. The GA fitness function only compares center-of-mass positions and velocities; the ring, warp, and stream matches are post-hoc inspection, not optimized targets. And only the single best COM-fitting orbit is analyzed, so we don't know how many other acceptable orbits also produce these features. The SMC position is off by 13.6 kpc, which is a lot, and the simulated ring's position angle differs from observed by ~60 degrees. Most importantly, the warp claim—the one in the abstract—does not survive across plausible LMC halo concentrations. Model F (more compact halo) suppresses the warp to small ripples; model G (more diffuse halo) gives an S-shape, not the observed U. The authors acknowledge this themselves in Section 5.10, but they still push the 140 Myr encounter as the cause of the U-shaped warp. That is a clear overreach without an argument that the real LMC halo falls in the window that produces a U-shape.

The paper's central scenario—recent, repeated close encounters—is plausible and consistent with several independent constraints, and the more robust results (ring, SMC expansion, distance bimodality) do not depend on the halo assumptions. The warp attribution specifically needs either a quantitative match metric or a demonstration that the halo concentration is constrained to the needed range. The lack of error bars on the encounter times is a lesser but real omission.

Who should read it: anyone actively working on Magellanic orbits or the origin of the LMC ring and warp. It deserves a serious referee: send it out, but the revision should quantify the warp match, run the feature comparison across the other GA solutions, and soften or defend the causal warp claim. I'd bring it to a reading group and probably cite the simulation and data release.

Referee Report

4 major / 5 minor

Summary. The paper uses a large suite of N-body simulations (pkdgrav) and a genetic algorithm to reconstruct the past 2.5 Gyr of interactions among the Milky Way, LMC, and SMC. The GA varies the initial positions and velocities of the two Clouds, running ~8,000 low-resolution simulations, and identifies a best-fit model that places the Clouds near their observed center-of-mass positions and velocities at the present day. This model has two close LMC-SMC encounters, ~940 Myr and ~140 Myr ago, and the authors claim it reproduces several observed stellar substructures: an LMC ring, the LMC northern stream, a U-shaped LMC warp, SMC tidal expansion, and an eastern SMC distance bimodality. The paper attributes the LMC warp to the most recent SMC encounter, which excites vertical oscillations with a mean amplitude of 1.3 kpc. The authors also vary LMC disc/halo parameters (models A-G) to test robustness; they find that a more compact halo (model F) suppresses the warp and a more diffuse halo (model G) produces an S-shaped warp.

Significance. If the reconstructed interaction history is correct, the paper would provide a unified dynamical explanation for several observed Magellanic substructures and would pin down the recent encounter times between the Clouds. The ambitious use of a genetic algorithm with live N-body galaxies is a methodological step forward, and the public release of the fiducial simulation is a valuable resource. The paper also makes a falsifiable prediction that vertical oscillations in the outer LMC should be observable. However, the strength of the claims is undercut by the qualitative nature of several feature matches, the large residuals in the GA fit, and the strong dependence of the central warp claim on an uncertain halo parameter. The paper is honest about its limitations but does not always carry that caution into the abstract and conclusions.

major comments (4)

[Sec. 5.10 / Abstract] The central claim that the ~140 Myr SMC encounter caused the observed U-shaped LMC warp is not robust to the assumed LMC dark-matter halo concentration. The paper shows (Sec. 5.10) that model F (R_h=13 kpc) produces only small-scale ripples with no coherent warp, while model G (R_h=32 kpc) produces an S-shaped warp rather than the observed U-shape. Since the halo concentration is not tightly constrained, the fiducial R_h=21.4 kpc result is just one plausible outcome. The authors themselves state, 'The range of parameter values we looked at left much uncertainty in the characteristics of the LMC.' Yet the abstract and Sec. 7 present the warp as a robust consequence of the interaction. This needs to be either softened to a conditional claim or supported by a quantitative warp comparison (amplitude, radial shape, U-shape parameter) across a range of halo concentrations to show that the U-sh
[Sec. 5.2] The simulated LMC ring is described as having a position angle of ~30° West of North (or ~150° East of North), whereas the observed ring has a position angle of roughly 20–30° East of North. This is a ~60° difference in orientation, yet the text states it is 'similar to what is seen in the simulation.' A 60° position-angle discrepancy is not a reproduction of the feature. The authors should either quantify the orientation uncertainty and show that projection effects or model choices can reconcile the difference, or present the ring match as partial and discuss what this discrepancy implies for the interaction geometry.
[Sec. 5.1 / Eq. (1)] The genetic algorithm fitness (Eq. 1) only uses the center-of-mass positions and velocities of the LMC and SMC. The final residuals are sizable: the SMC position is offset by 13.6 kpc (roughly 30% of its distance) and its velocity by 13.2 km/s. Because the substructure features (warp, ring, stream) were not part of the fitness function, their matches are post-hoc by construction. Moreover, the large SMC position offset means the derived interaction geometry—impact parameter and encounter times—is not tightly constrained. The paper should assess how the range of acceptable GA solutions (e.g., finalists or a nearby region of parameter space) changes the interaction times (940 Myr, 140 Myr) and whether the substructure matches persist. Without this, the specific encounter timing is not shown to be a robust result.
[Secs. 5.4, 5.5] The comparisons to observed features are largely qualitative. For the northern stream, Fig. 16 shows that the simulated in-plane radial velocity is lower in magnitude than the observed for much of the arm and even changes sign at the end; the authors call the trend 'similar' but offer no quantitative metric (e.g., rms difference, correlation). For the warp, Figs. 17–18 demonstrate vertical oscillations but there is no quantitative comparison of warp amplitude, radial extent, or U-shape against the Choi et al. (2018a) and Saroon & Subramanian (2022) measurements. Given the strong model dependence shown in Sec. 5.10, quantitative metrics are essential to support the claim that the model 'reproduces' the observed warp. I recommend adding such metrics for both features.

minor comments (5)

[Sec. 5.5] The text says 'At 140 Myr after the interaction the warp is above the disc plane... while at 300 Myr after the interaction it is below,' but the current snapshot is defined when the LMC crosses L_MS=0, which may not be exactly 140 Myr after the interaction. Please clarify the exact time offset between the current snapshot and the last encounter.
[Sec. 5.6] The simulated SMC rotation reaches ~38 km/s at 6 kpc after the interaction, while Zivick et al. (2021) find a maximum of ~20 km/s at 1 kpc. The paper notes this factor-of-two discrepancy but does not discuss its possible cause (e.g., mass model, stripping efficiency). A brief comment would help.
[Sec. 5.10] The statement that 'in all models, there are two interactions that occur at about the same times' is complicated by the fact that models B, C, and G cross L_MS=0 20–80 Myr earlier than model A, so the 'current' snapshot is not at the same physical time. Please clarify how this affects the comparison.
[Fig. 15 caption] Typo: 'Magellenic Stream' should be 'Magellanic Stream.'
[Fig. 28] The per-panel model labels (A, B, C, D, F, G) are small and partially obscured; larger labels would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the fitted quantities are COM positions/velocities; the reported substructures and interaction times emerge from the N-body evolution.

full rationale

The derivation chain is self-contained and non-circular. The genetic algorithm fitness (Eq. 1) is explicitly based on the final LMC/SMC median positions and velocities compared to observed values, and Section 4.1 states that adding density/radial velocity/proper-motion map comparisons 'did not have much of an effect' on the fits. The claimed outputs — the two close encounters at 940 Myr and 140 Myr, the LMC ring, the northern stream, the SMC tidal expansion, the eastern distance bimodality, and the LMC warp — are emergent properties of the evolved N-body models, not quantities entered into the fitness function. The paper reports offsets (LMC position 2.8 kpc, velocity 14.8 km/s; SMC position 13.6 kpc, velocity 13.2 km/s) rather than claiming exact reproduction of the fitted targets, further showing the outputs are not trivially forced. The LMC and SMC masses and halo parameters are adopted from external dynamical studies (Erkal et al. 2019; Besla et al. 2012) and are not derived from the features the paper claims to reproduce. The paper's own admission that the warp direction/amplitude differs across models F and G (Section 5.10) is a robustness limitation, not a circularity: it shows the causal claim depends on halo concentration, but the interaction itself exists in all models and the warp is not fitted. Self-citations such as Nidever et al. (2017, 2020) are observational data papers used for comparison, not load-bearing theoretical claims. No equation or parameter is defined in terms of the target results, no prediction reduces to a fitted input, and no uniqueness theorem or ansatz is imported from the authors' prior work. Therefore the paper's central derivation is independent of its own conclusions, and the correct circularity finding is a score of 0.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The model's orbit is shaped by fitted initial conditions and by adopted galaxy masses/halos from prior work; the feature matches are emergent but not independent of the fitted orbit. No new particles or forces are introduced.

free parameters (2)

Initial positions and velocities of LMC and SMC at t=-2.5 Gyr (12 numbers) = not quoted in paper; outputs of genetic algorithm
The GA varies x,y,z and vx,vy,vz for both galaxies to minimize the fitness in Eq. 1. These directly determine the close-encounter times (940 and 140 Myr ago) that anchor the central claim.
LMC stellar mass = 7.2 x 10^9 M_sun
Set by hand to the observed stellar mass plus gas and stripped mass (Section 3); raises the LMC's self-gravity and affects warp amplitude and tidal response.

assumptions (5)

domain assumption Observed COM positions and velocities in Table 2 are accurate and representative of the LMC and SMC centers.
Used as GA targets in Eq. 1; if the SMC velocity (Zivick et al. 2018) is biased, the fitted orbit and inferred encounter times change.
domain assumption The isolated-6-Gyr-then-interacting-2.5-Gyr setup with masses from Erkal et al. (2019) and Besla et al. (2012) captures the relevant physics.
Gas, star formation, and any pre-2.5 Gyr interactions are neglected; the authors state LMC hooks/SMCNOD may require earlier start (Sections 5.8 and 6).
ad hoc to paper The GA search over ~8000 simulations approximates the global optimum of the fitness landscape.
No convergence or multi-modal analysis is shown; the best solution retains a 13.6 kpc SMC position offset, so the orbit is not tightly constrained.
ad hoc to paper The LMC disc/halo models A-G bracket the real LMC, and the fiducial model A is the correct representation for drawing conclusions.
Model F removes the warp and model G changes its shape; the conclusion that the SMC encounter causes the observed U-shaped warp relies on model A.
domain assumption pkdgrav simulations at the stated resolutions resolve the tidal features being compared.
Orbit differences between MW resolutions are less than 2 kpc, but no convergence test is shown for warp/ring amplitudes.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Modeling the recent interactions between the Magellanic Clouds and Milky Way." pith.science (2026). https://pith.science/paper/F3OF6MXB

@misc{pith2026260205021,
  author       = {Pith},
  title        = {Pith review of: Modeling the recent interactions between the Magellanic Clouds and Milky Way},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/F3OF6MXB}},
  note         = {Machine review of arXiv:2602.05021}
}

read the original abstract

The Large and Small Magellanic Clouds (LMC and SMC, respectively) are the largest satellite galaxies of the Milky Way (MW) and their interactions with each other have given rise to multiple stellar substructures in their periphery as well as the gaseous Magellanic Stream. To better understand the origin of the stellar substructures and constrain their past orbit, we model the past 2.5 Gyr of the interactions between the MW and the LMC and SMC using N-body simulations. Due to the strong interactions, analytical orbit integrations are insufficient to analyze the past galaxy orbits accurately. Therefore, we use a genetic algorithm in combination with N-body simulations to determine the LMC and SMC initial positions and velocities 2.5 Gyr ago that result in the Magellanic Clouds (MCs) arriving near their observed locations and velocities at the current time. After running ~8,000 simulations, our best matching model includes two close interactions between the MCs (940 Myr and 140 Myr ago) and reproduces some observed features of the MCs, including the LMC disc warp, a ring-shaped overdensity in the LMC, the tidal expansion of the SMC, and a greater distance dispersion on the eastern side of the SMC. The LMC disc warp is caused by the most recent interaction with the SMC, which occurred ~140 Myr before the present. The interaction causes global ripples in the LMC disc with a mean amplitude of 1.3 kpc.

Figures

Figures reproduced from arXiv: 2602.05021 by the authors.

**Figure 2.** Density map of Gaia MC giant stars. The black line shows the selection we used on the map of the stars on Magellenic Stream coordinates 𝐿𝑀𝑆 and 𝐵𝑀𝑆. and MW. After Besla et al. (2007) discovered that the MCs are very likely on their first infall into the MW, a newer generation of simulations by Besla et al. (2012) and Diaz & Bekki (2012) explored the interactions between the LMC and SMC and were able to reproduce k… view at source ↗

**Figure 3.** Maps of the observed Gaia and SDSS MC data. left: Number of stars, center left: mean proper motion in the 𝐿MS direction, center right: mean proper motion in the 𝐵MS direction, and right mean line-of-sight velocity. 2 DATA The observational data we use to generate the MC proper motion and density maps come from Gaia EDR3 (Gaia Collaboration et al. 2021a), while the radial velocity measurements come from SDSS-IV / APO… view at source ↗

**Figure 4.** Diagram of the coordinates we define for our two-galaxy simulations. Top: projection onto the 𝑥 − 𝑦 plane. Bottom: projection onto the plane perpendicular to the 𝑥 − 𝑦 plane that includes the initial location of the SMC center and point P. parameter space by mimicking natural selection, evolving a population of candidate solutions through operations such as mutation, crossover, and selection to efficiently converg… view at source ↗

Figures from the paper (22 more)

**Figure 5.** Figure 5: The LMC and SMC at the end of a two-galaxy simulation. it were moving in a straight line, 𝑑 is the distance from the SMC center to P, 𝑣 is the initial speed of the SMC, 𝛽 is the angle from the 𝑥-axis counterclockwise to P, 𝛿 is the angle counterclockwise of the project…

**Figure 6.** Figure 6: The potentials we use in Gala compared to the potentials of the pkdgrav initial conditions. 50 0 50 100 x 0 50 100 150 200 y 100 50 0 50 100 x 50 0 50 100 150 z 0 50 100 150 200 y 50 0 50 100 150 z gala MW gala LMC gala SMC pkdgrav MW pkdgrav LMC pkdgrav SMC [PITH_FUL…

**Figure 7.** Figure 7: A comparison of orbits with Gala and pkdgrav using the same initial conditions. The Gala orbits are shown as solid lines while the pkdgrav orbits are dashed lines. While the orbits are fairly close at the beginning, they diverge and in the Gala run, the LMC and SMC mer…

**Figure 8.** Figure 8: The distance between the LMC and SMC over time for the best match simulation. The green dots are at the closest approaches and the orange dot shows the time for the snapshot when the LMC crosses 𝐿MS = 0 ◦ . 5.2 LMC Ring In our simulation, there is a high density ring a…

**Figure 9.** Figure 9: Comparison of density maps in MS coordinates between the best match simulation (left) and the Gaia data (right). 10 0 10 10 5 0 5 10 1.0 0.5 0.0 0.5 1.0 z 10 0 10 10 5 0 5 10 10 0 10 2 10 4 N* 10 0 10 10 5 0 5 10 20 0 20 vz [PITH_FULL_IMAGE:figures/full_fig_p007_9.png]

**Figure 10.** Figure 10: A face-on view of the inner LMC disc, showing the ring and colored by left: density, center: mean 𝑧, and right: mean 𝑉𝑧 . 0 2 4 6 8 R (kpc) 10 7 10 8 10 9 D e nsity (M /k pc 2 ) 0< <60 60< <120 120< <180 180< <240 240< <300 300< <360 [PITH_FULL_IMAGE:figures/full_fig…

**Figure 12.** Figure 12: The face-on map of the LMC stars with a deviation value, which is determined by how the velocity components differ from the median for their radius, greater than 20. 5.3 LMC kinematic substructure There are several signatures of LMC kinematic substructure in the simul…

**Figure 13.** Figure 13: Plots of 𝑣𝜙 vs 𝑣𝑅 for three different angle bins in the LMC (left: 30◦–60◦ , center: 60◦–90◦ , right: 90◦–120◦ ). Circles show an overdensity that changes location on the plot for different angles on the LMC. The bottom row shows a zoomed in version. bins out to a rad…

**Figure 14.** Figure 14: The combined stars of a kinematic feature of the LMC found through two different methods. from plots of the three velocity components as a function of LMC position angle, 𝜙. The northern arm begins at the northwest corner of the LMC and extends to the northeast corner…

**Figure 15.** Figure 15: The LMC plotted on Magellenic Stream coordinates in simulations which exclude (left) and include (right) the SMC. 0 5 10 15 20 25 Angle along arm (deg) 60 40 20 0 20 vR (k m s 1 ) Simulation Cullinane et al. (2022) [PITH_FULL_IMAGE:figures/full_fig_p009_15.png]

**Figure 16.** Figure 16: The in-plane radial velocity at different points along the length of our simulated northern arm compared to that measured by Cullinane et al. (2022). arm to observations, we plot the LMC on the sky, and measure the distance in degrees along the arm from where it separ…

**Figure 18.** Figure 18: Vertical oscillations in the LMC disc. Mean z values versus X in a narrow 2 kpc Y-slit for the 340 Myr after the most recent LMC-SMC collision. The vertical oscillations are clearly seen in both panels. 40 30 20 10 0 10 30 20 10 0 10 20 30 150 100 50 0 50 100 150 M e …

**Figure 19.** Figure 19: Face-on map of the SMC colored by mean 𝑣𝑅. map ( [PITH_FULL_IMAGE:figures/full_fig_p010_19.png]

**Figure 20.** Figure 20: Kinematics of SMC star particles. (Left) 𝑣𝑅 versus 𝑅. The filled circles show the median velocity value for each of the first six 1 kpc radial bins. A linear fit of 9.86 km s−1 kpc−1 is shown by the blue line. (Right) 𝑣𝜙 vs 𝑅. The median for each 1 kpc radial bin is a…

**Figure 21.** Figure 21: The tidal expansion of the SMC. The coordinate system here is centered on the LMC and the SMC is on the x-axis, so 𝑣𝑥 is the component of the SMC’s velocity toward or away from the LMC (negative is towards the LMC). The blue line shows the tidal expansion of 10 km s−1…

**Figure 22.** Figure 22: SMC 𝑣𝑅 vs. 𝑅 with different axis limits than [PITH_FULL_IMAGE:figures/full_fig_p011_22.png]

**Figure 23.** Figure 23: The face-on SMC in the SMC-centric coordinate system. Left: The stars below the cut in [PITH_FULL_IMAGE:figures/full_fig_p012_23.png]

**Figure 24.** Figure 24: The radial density of the SMC (left) in the simulation with two exponential discs (dashed lines) and their sum (dotted line) (right) from Gaia data. 30 20 10 0 10 20 30 LMS 30 20 10 0 10 20 30 B MS 10 0 10 1 10 2 10 3 N* [PITH_FULL_IMAGE:figures/full_fig_p012_24.png]

**Figure 25.** Figure 25: Map of simulated SMC stars on Magellanic Stream coordinates with dashed lines dividing the SMC into different regions. models. The SMC is also in a similar location south of the LMC in model F. In the others, the SMC is in a similar location in the sky as for the fidu…

**Figure 26.** Figure 26: Histograms of distances to simulated SMC stars. Each histogram is of stars in the box in the same grid location as in [PITH_FULL_IMAGE:figures/full_fig_p013_26.png]

**Figure 27.** Figure 27: Density map on the sky and distance vs. 𝐵𝑀𝑆 for simulated SMC stars. The left panels include all SMC stars, the center panels include only the stars where 𝑣𝜙 > 0, and the right panels include only the stars where 𝑣𝜙 < 0. 20 0 20 B MS A B C 20 0 20 LMS 20 0 20 B MS D 2…

**Figure 28.** Figure 28: The maps of the different simulations at the ‘current’ snapshot for each model. Top row: models A, B, and C. Bottom row: models D, F, and G. MNRAS 000, 1–16 (2026) [PITH_FULL_IMAGE:figures/full_fig_p014_28.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

LMC-induced Perturbations in the Milky Way Halo II: Bridging Field-level Inference and Summary-level Simulation-Based Inference
astro-ph.GA 2026-07 conditional novelty 6.0 of 10

A field-level flow-matching likelihood shows the raw 6D halo phase-space distribution carries 2.5-9.9x more MW-LMC parameter information than velocity moments; adding BFE+MOPED summaries recovers much of this gap.

Reference graph

Works this paper leans on

1 extracted references · 1 linked inside Pith · cited by 1 Pith paper

[1]

A., Haffner L

Abdurro’uf et al., 2022, ApJS, 259, 35 Almeida A., et al., 2024, MNRAS, 529, 3858 Andrae R., Rix H.-W., Chandra V., 2023, ApJS, 267, 8 Barger K. A., Haffner L. M., Bland-Hawthorn J., 2013, ApJ, 771, 132 Belokurov V. A., Erkal D., 2019, MNRAS, 482, L9 Besla G., 2015, arXiv e-prints, p. arXiv:1511.03346 BeslaG.,KallivayalilN.,HernquistL.,RobertsonB.,CoxT.J....

arXiv 2022

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [1]

A., Haffner L

Abdurro’uf et al., 2022, ApJS, 259, 35 Almeida A., et al., 2024, MNRAS, 529, 3858 Andrae R., Rix H.-W., Chandra V., 2023, ApJS, 267, 8 Barger K. A., Haffner L. M., Bland-Hawthorn J., 2013, ApJ, 771, 132 Belokurov V. A., Erkal D., 2019, MNRAS, 482, L9 Besla G., 2015, arXiv e-prints, p. arXiv:1511.03346 BeslaG.,KallivayalilN.,HernquistL.,RobertsonB.,CoxT.J....

arXiv 2022