A Galactic Transformation -- Understanding the SMC's Structural and Kinematic Disequilibrium

TL;DR

The paper tackles the longstanding puzzle of the SMC's disequilibrium by invoking a recent direct collision with the LMC as the primary driver of its distorted morphology and kinematics. Using hydrodynamic N-body simulations (Model 2) and a non-equilibrium analysis pipeline, it demonstrates that tidal tails create the large line-of-sight depth, and that stellar kinematics become dispersion-dominated while gas dynamics are governed by radially outward flows, partly due to ram-pressure during the collision. The work shows that the observed HI velocity gradient does not reflect disk rotation but radial gas motions, and that gas–stellar center offsets naturally arise from the collision-driven perturbations. It also highlights the limitations of equilibrium mass estimators (e.g., Virial) for the SMC and argues for non-equilibrium approaches, including leveraging perturbations in the LMC disk, to constrain the SMC's dark matter content. Overall, the findings suggest that group processing via close galaxy collisions can drive rapid dIrr to dE/dSph transformation and efficient gas removal, with broad implications for interpreting low-mass galaxy dynamics and ISM physics.

Abstract

The SMC is in disequilibrium. Gas line-of-sight (LoS) velocity maps show a gradient of $60-100$ km s$^{-1}$, generally interpreted as a rotating gas disk consistent with the Tully-Fisher relation. Yet, the stars don't show rotation. Despite a small on-sky extent ($\sim4$ kpc), the SMC exhibits a large ($\sim10$ kpc) LoS depth, and the stellar photometric center is offset from the HI kinematic center by $\sim$1 kpc. With N-body hydrodynamical simulations, we show that a recent ($\sim$100 Myr ago) SMC-LMC collision (impact parameter $\sim2$ kpc) explains the observed SMC's internal structure and kinematics. The simulated SMC is initialized with rotating stellar and gaseous disks. Post-collision, the SMC's tidal tail accounts for the large LoS depth. The SMC's stellar kinematics become dispersion dominated ($v/σ\approx0.2$), with radially outward motions at $R>2$ kpc, and a small ($<10$ km s$^{-1}$) remnant rotation at $R<2$ kpc, consistent with observations. Post-collision gas kinematics are also dominated by radially outward motions, without remnant rotation. Hence, the observed SMC's gas LoS velocity gradient is due to radial motions as opposed to disk rotation. Ram pressure from the LMC's gas disk during the collision imparts $\approx30$ km s$^{-1}$ kick to the SMC's gas, sufficient to destroy gas rotation and offset the SMC's stellar and gas centers. Our work highlights the critical role of group processing through galaxy collisions in driving dIrr to dE/dSph transformation, including the removal of gas. Consequently, frameworks that treat the SMC as a galaxy in transformation are required to effectively use its observational data to constrain interstellar medium and dark matter physics.

Figures (15)

• Figure 1: Left panel: Placing the SMC on the Baryonic Tully-Fisher Relation (BTFR). The black dashed line denotes the BTFR fit taken from McGaugh2000, with the grey shaded region denoting the $1-\sigma$ uncertainty. The SMC's total baryonic mass (stars $+$ HI) is $(7.0 \pm 0.6) \times 10^8$ M$_\odot$Stanimirovic1999Harris2004. Symbols denote the SMC's inferred peak HI rotation velocity of $56 \pm 5$ km s$^{-1}$DiTeodoro2019 and the peak old star rotation velocity of $< 10$ km s$^{-1}$Zivick2021. The kinematics of the SMC's HI and old stars are strongly discrepant. The orange circle denotes the SMC's initial disk in the B12 simulation, which is consistent with BTFR by design. Right panel: the SMC's photometric center $(12.80^\circ, -72.83^\circ)$Gonidakis2009 and the HI kinematic center $(16.26^\circ, -72.42^\circ)$Stanimirovic2004 are plotted over a background of SMC stars selected from the NN Optimal Gaia DR3 sample of Arranz2023. The two centers are separated by $1-2$ kpc on-sky, which is a significant fraction of the SMC's on-sky extent of $\approx 4-5$ kpc. Explaining the kinematic discrepancies between the old stars and HI (rotation peaks and centers) is the goal of this work.
• Figure 2: The SMC's orbit about the LMC in B12 Model 1 (orange dash-dot line) and Model 2 (purple solid line) simulations, after their MW infall. Three additional key epochs corresponding to Model 2 are marked with vertical dashed lines: SMC-SMC collision (impact parameter $\approx 2$ kpc); 100 Myr and 200 Myr post-collision. The fiducial present day is denoted as time $= 0$, and in Model 2, this corresponds to 100 Myr post-collision. The Model 2 SMC is expected to be highly morphologically and kinematically disturbed post-collision, and will be the primary subject of investigation in this work. Model 1, without a collision, serves as a control.
• Figure 3: Top row: Identifying the stellar density center of the simulated SMC. The left (right) panel shows the x-y (x-z) projection of the SMC's stellar surface density distribution 100 Myr after the SMC-LMC collision in B12 Model 2. The Galactocentric axes are translated to the inferred density center, depicted by the red star at (0, 0). The red star is a reasonable representation of the SMC's stellar density peak (a $10^7$ M$_\odot$ kpc$^{-2}$ contour is included to guide the eye). The iterative shrinking sphere method succeeds in identifying the SMC's stellar density center despite its highly disturbed morphology. Bottom row: Computing the systemic velocity of the simulated SMC. The left (right) panel shows the $v_{x}$ - $v_{y}$ ($v_{x}$ - $v_{z}$) projection of the SMC's stellar velocity distribution for the same epoch as the top row. The Galactocentric velocities are translated to the calculated stellar systemic velocity, depicted by the cyan star at (0, 0). The colorscale depicts star particle counts in 4 km s$^{-1}$ by 4 km s$^{-1}$ bins. The cyan star is a reasonable representation of the center of the SMC's velocity field (a contour corresponding to 75% of the peak value of the distribution is shown to guide the eye). The iterative shrinking sphere method succeeds in identifying the SMC's systemic velocity despite its disturbed kinematics.
• Figure 4: Identifying the stellar kinematic center of the Model 2 simulated SMC. The SMC is oriented edge-on ($x' - z'$ projection) and the in-plane velocities are mapped ($v_y'$ in this projection) through the color bar. The red (blue) colors depict the values of $v_y'$ for stars going into (coming out of) the plane of the paper. Three epochs are shown - MW infall ( left panel), 100 Myr, and 200 Myr post SMC-LMC collision ( middle panel and right panel respectively). The SMC's stellar kinematic center (translated to the origin) is marked by the black star. The black star is a reasonable representation of the center of the SMC's stellar rotation field, even when the internal kinematics are significantly disturbed and the amplitude of rotation is small ($< 20$ km s$^{-1}$).
• Figure 5: Time evolution of the Model 2 SMC's stellar surface density distribution in the plane of rotation. MW infall epoch, 100 Myr and 200 Myr post-collision are shown. The red star marks the stellar density center. Contour levels represent 5%, 10%, 20%, 40% of the peak surface density. The arrow points towards the LMC. Post-collision, the SMC's stellar distribution becomes significantly elongated along tidal structures. Bottom right: the spherically averaged stellar density profile for the SMC at different epochs. At MW Infall/Model 1 control, the SMC stellar distribution is roughly consistent with an exponential disk. Post-collision, the stellar density in the SMC's interior (R $\lesssim 3$ kpc) decreases by a factor of 2 - 3, and that in the SMC's outskirts (R $\gtrsim 3$ kpc) increases by a factor of 2 - 3, relative to the infall epoch. Post-collision stellar density profile is a combination of at least two power laws, with a break around 2 - 4 kpc. This means that the LMC's tidal influence is significant within 3-4 scale lengths of the initial SMC disk.