Michele Fumagalli

Highlights

Ordinary galaxies wear their metals farther out than their starlight

Median-stacked images of Mg II (top) and [O II] (bottom). Left: the line emission. Middle: the starlight in control windows. The line emission is visibly broader than the stars that produced it. Dutta et al. 2023, MNRAS.
Median-stacked images of Mg II (top) and [O II] (bottom). Left: the line emission. Middle: the starlight in control windows. The line emission is visibly broader than the stars that produced it. Dutta et al. 2023, MNRAS.

Almost everything we know about the gas around galaxies, we know from shadows. A quasar sits behind a galaxy, its light passes through the halo, and the gas writes itself into the spectrum as absorption. It is a superb technique with one structural limitation: each quasar gives you a single pencil-thin line of sight. You learn what the gas is like at exactly one point, and nothing at all about its shape.

Seeing the gas glow instead would fix that. The difficulty is that it barely does — and the metals in it, which are what tell you where the gas has been in the past, glow more faintly still.

So we stopped looking at individual galaxies. Taking roughly six hundred galaxies from two deep MUSE surveys, at redshifts between 0.7 and 1.5, we stacked them on top of one another until the average halo emerged from the noise. Magnesium and oxygen both appear, and both extend farther than the starlight that produced them — on average out to about 25 and 45 kiloparsecs respectively, at a surface brightness of 10⁻²⁰ erg s⁻¹ cm⁻² arcsec⁻². That is the first such measurement for a general population rather than for the extreme, spectacularly bright systems that dominate the literature.

The two lines behave differently, and the difference is informative. Magnesium falls off more gently with radius than oxygen does — expected, because magnesium is a resonant line, scattered and reprocessed on its way out, and sensitive to dust. Meanwhile the ratio of oxygen to magnesium holds at about three across 20–40 kiloparsecs, which is hard to explain unless a good deal of this light is being produced in place in the halo rather than leaking out from the galaxy.

Two trends fall out of the stacks. The metal-enriched gas reaches farther around more massive galaxies, and farther again around galaxies sitting in groups — the environments where interactions strip gas out of galaxies and leave it in the space between them. And the haloes are brighter and more extended at the higher end of our redshift range than the lower, back when these galaxies were forming stars more vigorously.