QED V: Variations in metal loading of galactic winds with element nucleosynthetic origin

TL;DR

This study demonstrates that metals returned by different stellar nucleosynthetic channels are not equally loaded into galactic winds. Using high-resolution tall-box ISM simulations that separately track Type II SNe, Type Ia SNe, and AGB ejecta, the authors show substantial, environment-dependent differential metal loading, with typical variations of about $0.3$ dex in the fraction of newly synthesized metals lost to winds. A key finding is that the differential loading correlates with the relative volume-filling of supernova remnants, and that the resulting abundance shifts can mimic or obscure interpretations of star-formation histories and IMFs derived from chemical diagnostics. These results call for caution when using abundance ratios in galaxy formation studies and motivate broader, cosmological zoom-in simulations to map the dependence of differential loading on galaxy properties.

Abstract

Type Ia supernovae, type II supernovae, and asymptotic giant branch (AGB) stars are important sites of stellar nucleosynthesis, but they differ greatly in their rates, their location within a galaxy, and the mean thermal energy and abundance distribution of their ejecta. In earlier papers in this series we have shown that a significant fraction of metals newly synthesized by type II supernovae are promptly lost to galactic winds -- i.e., galactic winds are metal loaded. Here we investigate whether the elements returned by type Ia supernovae and AGB stars are similarly metal loaded, or whether metal loading varies significantly with nucleosynthetic site. We use a series of high-resolution ``tall box'' simulations of the interstellar medium with the \quokka~GPU-accelerated code, within which we systematically vary the galaxy gas surface density, metallicity, and the scale heights and relative rates of the different nucleosynthetic sources. We show that the metal loadings of galactic winds differ substantially between metals produced by different sources, with typical variations at the level of $\approx 0.3$ dex, a phenomenon we term differential metal loading. Which set of metals suffers preferential loss from this phenomenon varies depending on the galactic environment, and is not easily predictable \textit{a priori}. Our findings call into question the the interpretation of diagnostics of galaxy formation, for example star formation timescales and initial mass functions, based on abundance diagnostics, since the abundance variations upon which these techniques rely are often at levels comparable to those we show can be induced by differential metal loading.

Figures (6)

• Figure 1: A slice through the plane $y=0$ for run $\Sigma$6.6 at $t=60$ Myr. From (a) to (f), the quantities shown are gas density, temperature, SN type II metal abundance normalised to Solar, the metal mixing ratios (\ref{['eqn:colour']}) for type II, type Ia, and the AGB. Column (g) shows the colour with the red green, and blue channels mapped to the mixing ratios of type II, type Ia, and AGB ejecta, respectively -- see \ref{['ssec:colour']} for details. Note that, for clarity, we show only the upper half of the domain; the full simulation domain extends to $z=-4$ kpc.
• Figure 2: Metal loading factors $\eta_{Z,k}$ (\ref{['eq:etaZ']}) and $\phi_k$ (\ref{['eq:phi']}) for the $\Sigma$6.6 run. In both panels solid lines and points show time averages, and shaded bands show the $16$th to $84$th percentile variation in time over the full run. Red, green, and blue colours correspond to the metal fields tracing type II ejecta, type Ia ejecta, and AGB ejecta, respectively.
• Figure 3: Slices through run $\Sigma$6.6 at the same time as shown in \ref{['fig:slice_sig4']}. Each row shows a slice in the $xy$ plane at a height $z$ indicated in the legend. The left column shows colour (as in the right panel of \ref{['fig:slice_sig4']} -- see \ref{['ssec:colour']}), while the middle column shows vertical velocity $v_z$. The right column shows the distribution of metal mass with respect to velocity for each of the three metals -- $Z_\mathrm{II}$ (red dot-dotted), $Z_{\rm Ia}$ (green dashed), and $Z_\mathrm{AGB}$ (blue solid) -- in the slice.
• Figure 4: Time-averaged mixing ratio (\ref{['eqn:colour']}) as a function of height for gas that is hot ($T > 5\times 10^5$ K; left) and warm ($T < 2\times 10^4$ K; right) for type II (red), type Ia (green), and AGB (blue) ejecta. The quantities plotted are averages over all cells and over all times $t>50$ Myr, the approximate time at which the outflow reaches steady-state. Black dashed horizontal lines indicate $1/3$, the value corresponding to material that is fully mixed.
• Figure 5: Time-averaged metal loading factors $\eta_Z$ (top; \ref{['eq:etaZ']}) and $\phi$ (middle; \ref{['eq:phi']}), and differential metal retention $\epsilon$ (bottom; \ref{['eq:epsilon']}) for all metal types in all simulations.