Observations of a Magellanic Corona

The study investigates whether the Large Magellanic Cloud hosts a warm–hot, collisionally ionized circumgalactic corona as implied by its high halo mass (≈10^11.1–10^11.4 M☉). Such a corona, at roughly virial temperatures (10^5.3–10^5.5 K) and extending to ~100–130 kpc, would influence the formation and evolution of the Magellanic Stream and the baryon cycle in the Local Group. The research question is addressed using ultraviolet absorption-line spectroscopy of background quasars near the LMC to search for high-ionization species (C IV, Si IV, O VI) that would trace multiphase CGM gas and, specifically, a Magellanic Corona. The purpose is to detect and characterize this corona, measure its radial profile and mass, and assess its role within the Magellanic system and the Local Group’s hierarchical assembly. The study’s importance lies in closing the baryon budget of the Magellanic group, understanding gas accretion onto the Milky Way, and providing direct evidence for satellite-galaxy coronae.

Prior work established that the LMC and SMC are likely on their first infall toward the Milky Way and that the LMC’s halo mass is sufficiently large to support a warm–hot corona. The Magellanic Stream has been extensively mapped and studied as a major gaseous structure feeding the Milky Way with metal-poor gas. Surveys of the CGM around dwarf galaxies (e.g., COS-Dwarfs) and around M31 provided benchmarks for radial profiles of metal-line absorption. Possible contributors to high-ionization absorption along Magellanic sightlines include photoionization from a Galactic Seyfert-like flare and shock-cascade models in the Stream, but these mechanisms are restricted in geometry or do not reproduce the observed radial trends. Theoretical models of collisional ionization, including non-equilibrium cooling, and interface physics (turbulent/conductive mixing layers) predict characteristic ion ratios and temperatures for C IV, Si IV, and O VI that guide interpretation of the observations.

Data: The study analyzes 28 HST/COS far-UV spectra of background quasars within 45° of the LMC (impact parameter p_LMC < 35 kpc), each with S/N > 7 per resolution element. Six of these lines of sight have usable archival FUSE spectra to measure O VI λλ1031,1037. Instrumentation and spectral setup: COS G130M (≈1150–1450 Å) and G160M (≈1405–1775 Å) were used, providing spectral resolution FWHM ≈20 km s−1 (G130M) and ≈15 km s−1 (G160M) with native pixel size ≈2.5 km s−1; spectra were binned to Nyquist sampling. Lines analyzed include O I 1302, N I 1199–1200.7, C II 1334, Al II 1670, Si II multiplets (1260, 1193, 1190, 1526, 1304; 1250–1259), Fe II 1608,1144, Si III 1206, C IV 1548,1550, and Si IV 1393,1402. FUSE LIF1 channel covers O VI with ≈20 km s−1 resolution and 2 km s−1 native pixels; data were Nyquist binned. Night-only reductions mitigated geocoronal contamination (e.g., O I 1302, Si II 1304). Customized reductions and wavelength calibrations were applied. Sample selection and geometry: Impact parameters were computed using p_LMC = D sin(θ) with D = 50 kpc. Analysis was limited to θ ≤ 45° to ensure reliable p_LMC and separation from Milky Way absorption in velocity space. Absorption-line analysis: Spectra were continuum-normalized (third-order polynomials). Voigt profile fitting employed the VoigtFit Python package with recent atomic data and Gaussian instrumental LSF approximations. Low/intermediate ions (O I, N I, C II, C II*, Si II, Si III, Al II, Fe II) were fit jointly with component centers tied when appropriate. Saturated Si III required tying b-parameters to Si II. C II* contamination of C II at +250 km s−1 was handled by fixing the C II column to 10^13.8 cm−2 in blended cases, and those measurements were excluded from science analysis. High ions (C IV, Si IV) were fit jointly but independently of low ions to avoid bias; O VI was fit independently in FUSE when feasible. Upper limits were computed from S/N. Components associated with the Milky Way or known IVC/HVCs were flagged and excluded. Kinematic selection: Only components with |v_LSR| > 150 km s−1 were retained to isolate Magellanic gas, based on observed distributions and simulations that separate Galactic and Magellanic components; dynamical arguments place most Magellanic gas within 180–380 km s−1 around the LMC velocity. Ionization modeling: Low-ion absorbers were modeled as photoionized gas with derived typical properties log10(T/K) = 4.02 ± 0.07, log10(n_H/cm−3) = −1.4 ± 0.3, and line-of-sight sizes log10(L/kpc) = −1.5 ± 0.4. High ions (C IV, Si IV) were modeled using collisional ionization, both equilibrium and non-equilibrium isochoric scenarios, to interpret ion ratios (C IV/Si IV, O VI/C IV, O VI/Si IV). O VI is attributed to a hotter phase near 10^5.5 K where its collisional ionization fraction peaks. Photoionization was assessed and ruled out for the high-ion columns considered. Radial profiles and mass estimates: Column densities of C IV, Si IV, and O VI were examined versus p_LMC, revealing declining profiles. Linear regression fits (with 68% confidence intervals) quantified trends; inner sightlines (p_LMC < 7 kpc) show deficits in collisionally ionized high ions and were treated separately for some fits. Hydrogen column densities (N_H) for each phase were inferred from the ionization models. Metallicity assumptions: [Z/H] = −0.67 for photoionized gas; [Z/H] = −1 for interface/coronal high-ion gas. Integrating fitted N_H profiles within p_LMC < 35 kpc yielded ionized mass estimates for each phase. Photoionized results used Cloudy-based models; high-ion phases used CIE and non-equilibrium isochoric models.

• Detection and characterization of a Magellanic Corona: Evidence includes direct O VI absorption (O+5) and widespread C IV and Si IV absorption at Magellanic velocities, inconsistent with photoionization and requiring collisional ionization.
• Multiphase Magellanic CGM: At least three phases are detected: photoionized low-ion gas (~10^4 K), interface gas traced by C IV/Si IV (~10^5 K), and a hotter coronal phase traced by O VI (~10^5.5 K). C IV/Si IV ratios imply T ~ 10^5 K; O VI/C IV and O VI/Si IV require higher temperatures, indicating a distinct hotter phase.
• Radial decline: Significant anti-correlation of C IV and Si IV column densities with LMC impact parameter; O VI shows a similar but less certain decline due to small sample size. Inner sightlines (p_LMC < 7 kpc) show a relative deficit of collisionally ionized high ions compared to 7–12 kpc.
• Covering fraction: C IV covering fraction is 78% within 25 kpc and 30% for 25–35 kpc.
• Mass budget within 35 kpc: • Photoionized phase (~10^4 K): log10(M_HI,photo/M☉) = 8.7^{+0.2}_{−0.1} (assuming [Z/H] = −0.67). • Interface high-ion phase (~10^5 K; C IV, Si IV): log10(M_HI,EQ/M☉) = 8.5 ± 0.1 (equilibrium) and log10(M_HI,Non-Eq/M☉) ≈ 8.6 ± 0.1 (non-equilibrium) for [Z/H] = −1. If this gas were as cool as ~10^4.3 K in the non-equilibrium case, the mass would increase by ~1 dex. • Coronal O VI phase (~10^5.5 K): log10(M_HI/M☉) ≈ 8.3 ± 0.1 (equilibrium) and ≈ 8.5 ± 0.1 (non-equilibrium), assuming [Z/H] = −1. • Total ionized Magellanic CGM: log10(M_HI,LCGM/M☉) ≈ 9.1 ± 0.2 across phases.
• Kinematics and ionization: High-ion components are centered on LMC velocities and distinct from Milky Way gas. O VI traces a hotter phase with velocity offsets indicative of a coronal component; many sightlines showing high ions also show low ions, consistent with interfaces.
• Sample statistics: 28 sightlines analyzed; 52 unique Magellanic high-ion (C IV/Si IV) components and 10 O VI components (in 6 FUSE-detected sightlines) passed selection criteria.

The observations support a model in which most C IV and Si IV arise in turbulent or conductive interfaces between cooler ~10^4 K CGM clouds and a hotter ~10^5.5 K Magellanic Corona dominated by the LMC. This interface model explains the observed radial decline (more cool clouds and thus more interfaces at smaller radii), the kinematic alignment of high ions with low ions, the broader Si IV line widths relative to Si II, and the presence and velocity structure of O VI that directly traces the hotter corona. Alternative scenarios are disfavored: a diffuse ~10^4 K corona would be thermally unstable and does not account for O VI, and interfaces with a hot Milky Way corona cannot reproduce the observed dependence on distance from the LMC rather than the Milky Way. While the Magellanic Stream and possible Galactic Seyfert flare photoionization or shock cascades may contribute along specific sightlines, they do not explain the overall radial trends seen in both low and high ions. Comparisons to CGM profiles around low-mass galaxies (COS-Dwarfs) and M31 suggest a truncated or steeper profile around the LMC, though uncertainties in virial radii complicate direct comparisons. Despite some sightlines passing closer to the SMC, the more massive LMC likely dominates a single, enveloping Magellanic Corona. The inferred coronal electron columns suggest detectability via dispersion measures of extragalactic fast radio bursts intersecting the LMC halo, offering an independent probe. Overall, these findings indicate a nested, co-evolving ecosystem where the Magellanic group fell into the Milky Way with its own CGM, contributing substantially to the baryon budget and influencing gas accretion processes in the Local Group.

This work presents evidence for a Magellanic Corona around the LMC detected via UV absorption of O VI and high ions C IV and Si IV, revealing a pervasive, multiphase Magellanic CGM with a declining radial profile out to at least 35 kpc. Ionization modeling and kinematics indicate three phases—photoionized (~10^4 K), interface (~10^5 K), and coronal (~10^5.5 K)—with a combined ionized mass of log10(M_HI,LCGM/M☉) ≈ 9.1 ± 0.2. Among competing scenarios, interfaces between cool clouds and an LMC-dominated corona best explain the data. These results advance understanding of the Magellanic group’s gas reservoir, its role in the Local Group’s hierarchical assembly, and its contribution to the baryon budget. Future work should expand O VI samples and sky coverage, incorporate full 3D modeling to overcome projection effects beyond 45°, refine metallicity and non-equilibrium treatments, and leverage fast radio burst dispersion measures to independently quantify the coronal electron content.

• Sample and coverage: Only 28 sightlines within 45° (p_LMC < 35 kpc) were analyzed, with O VI measurable in just six sightlines, limiting statistical power for the hottest phase and radial extent beyond 35 kpc.
• Projection and geometry: Limiting to θ ≤ 45° avoids severe projection uncertainties; beyond this, a full 3D absorber–galaxy geometry would be required, and separation from Milky Way absorption becomes more difficult.
• Instrumental/systematic effects: Fixed pattern noise can persist after reduction and may affect measured columns; Gaussian LSF approximation may introduce minor biases; saturation (e.g., Si III) and C II* contamination required constraints or exclusions.
• Component selection: Velocity threshold |v_LSR| > 150 km s−1 reduces but may not eliminate Galactic contamination; association with known IVC/HVCs was excluded by flags but residual ambiguities may remain.
• Modeling assumptions: Metallicities were assumed ([Z/H] = −0.67 for photoionized, −1 for high ions); high-ion gas likely departs from equilibrium, affecting inferred temperatures and masses. Non-equilibrium isochoric models allow cooler solutions (~10^4.3 K) that would increase mass estimates by ~1 dex for interfaces.
• Comparisons to other halos: Uncertainties in virial radii complicate direct comparison of radial profiles with COS-Dwarfs and M31 surveys.