First statistical constraints on galactic scale outflows properties traced by their extended Mg II emission with MUSE

TL;DR

This paper provides the first population-level constraints on galactic-scale outflows traced by extended Mg II emission using deep MUSE data for 47 galaxies in the redshift range $0.7<z<2.0$. By applying a Sobolev-style outflow model to an ensemble of radially accelerating shells within a biconical geometry, the authors infer wind properties such as launching velocity $v_0$, terminal velocity $v_{\rm max}$ (up to ~$490$ km s$^{-1}$), and a near-linear velocity-radius relation ($\gamma\approx1$). The inner regions are optically thick ($\log\tau_0\approx2.6$) and halos have a typical half-light radius around $5$ kpc, with a tail extending to $\sim20$ kpc; halo sizes do not show a strong dependence on stellar mass, though compact halos ($\mathrm{HLR}<8$ kpc) may correlate with $M_*$. The study finds no universal alignment between outflow orientation and galaxy disks, and reveals that the prevalence and properties of Mg II halos depend on data depth and host properties, indicating a multifaceted link between galactic winds and the circumgalactic medium. These results advance understanding of cool-gas outflows and baryon cycling in galaxy evolution, while highlighting the need for larger samples and more sophisticated radiative-transfer models to fully capture CGM complexity.

Abstract

Galaxies evolve within vast gaseous halos that fuel star formation and carry signatures of feedback-driven outflows. Deep integral field data have enabled the study of MgII halos, which trace galaxy-scale outflows in emission, but their faintness has limited studies to single-object analyses. Here, we present the first statistical study of MgII-emitting halos using deep MUSE observations of 47 star-forming galaxies at $0.7<z<2.0$. Building on our previous work, where we developed and applied an outflow modeling framework for a single MgII halo, we now extend this approach to a larger sample, enabling robust population-level insights on the properties of circumgalactic outflows traced by their extended MgII emission. We detect extended emission out to tens of kiloparsecs and model the outflows as an ensemble of radially accelerating shells. Galaxies with MgII outflows tend to have higher SFRs, sSFRs, and younger stellar populations, consistent with star-formation-driven winds. The observations are consistent with winds that accelerate linearly with radius, from launching velocities of ~60 km/s up to maximum velocities that correlate with stellar mass and reach ~490 km/s. Their inner regions are highly opaque, and we find a tentative trend between stellar mass and central optical depth. The opening angle of the outflow shows some dependency on the host-galaxy stellar mass, with less massive galaxies showing primarily wide opening angles, and more massive galaxies showing a broader range of values, with both wide and narrow opening angles. The distribution of the spatial extent of MgII halos exhibits a clear peak at half-light radius (HLR) of ~5 kpc, with an extended tail of larger HLR values, up to ~20 kpc. Compact halo sizes (HLR $< 8$ kpc) correlate with stellar mass, but extended halos do not, which could suggest a difference in the powering mechanism between compact and extended halos.

Figures (36)

• Figure 1: Stellar mass distribution of our preliminary sample of galaxies. The sample consists of 89 galaxies drawn from the MUSCATEL survey that exhibit Mg II either in emission, absorption, or a P-Cygni profile, determined after a careful inspection of their integrated spectra and their continuum-subtracted Mg II pseudo-narrowband image. A direct S/N threshold has not been applied. Each stacked histogram shows the distribution of the galaxies that belong to each one of the depth levels of the MUSCATEL survey (shallow field, medium field, and deep field, see Sec. \ref{['sec:obs']}), indicated in the top-left part of each panel. The black line shows the distribution of the full preliminary sample in all the panels. Stellar masses are calculated via SED fitting, as described in Sec. \ref{['sec:photometry']}. The stacked histograms are color-coded by the different spectral profiles shown in Mg II, as indicated in the legend.
• Figure 2: Significance of the extended Mg II emission, stellar mass, and redshift distribution for the galaxies in our preliminary sample that exhibit a P-Cygni profile in Mg II. Left: Significance of the extended Mg II emission, in terms of its spatially integrated S/N, for the 50 galaxies in our preliminary sample that exhibit a P-Cygni profile in Mg II. We further removed from our preliminary sample those galaxies where the integrated S/N of the modeled emission is lower than $3\,\sigma$ (indicated with a vertical dashed red line). Middle: Stacked histogram that shows the stellar mass distribution of galaxies in the preliminary sample with a P-Cygni profile in Mg II. The gray histogram shows the stellar mass distribution of those galaxies where the significance of the Mg II emission is above the $3\,\sigma$ threshold. The black histogram shows the stellar mass distribution of galaxies below this significance threshold. Right: Same as middle panel, for the redshift distribution of the sample galaxies.
• Figure 3: Star formation rate as a function of stellar mass, for our sample galaxies. Both quantities are obtained via SED fitting (see Sec. \ref{['sec:photometry']}). Our final sample of 47 galaxies that exhibit a P-Cygni profile and a significant detection of modeled extended emission is shown with squares, color-coded by redshift. The preliminary sample is shown as dark gray circles, and the original parent sample is shown as smaller light gray dots. For reference, we show the SFMS as measured by Popesso2023 for $z=1$ (solid) and $z=2$ (dashed), by collecting different measurements of the SFMS for a wide range of redshifts.
• Figure 4: Compilation of continuum-subtracted Mg II pseudo-narrowband images (gray color scale) and HST F814W images (blue color scale, probing the stellar light) for our final sample of galaxies, that exhibit a P-Cygni profile in Mg II, as well as a significant detection of extended Mg II emission. Galaxies are sorted by stellar mass in descending order. The continuum-subtracted Mg II pseudo-narrowband images are computed by collapsing the MUSE data across the wavelength axis, for wavelengths that enclose the full P-Cygni profile of the Mg II doublet, for a FoV of 80 x 80 kpc$^{2}$. The brown circle in the bottom left part of the panel shows the size of the MUSE PSF. The black contours show the $1-\sigma$ and $2-\sigma$ detection levels of Mg II net emission in the pseudo-narrowband images. The red contours show the $2-\sigma$ and $4-\sigma$ levels of Mg II net absorption in the pseudo-narrowband images. For visualization purposes, the continuum-subtracted Mg II pseudo-narrowband images have been smoothed using a Gaussian kernel with a full width at half maximum of 1 arcsec. For each galaxy, we also show the radial profile of the Mg II emission as observed in the data (solid) and after correcting by self-absorption using our best-fitting model (dashed, see Sec. \ref{['sec:reconstruction']}). The vertical brown dotted line shows the half-light radius measured for the self-absorption corrected radial profiles. The inset shows the Mg II spectrum extracted from the MUSE data cube for each galaxy, across the full modeled region (black) and a small central aperture of radius $0\farcs4$.
• Figure 4: (continued)