Enhanced rates of stellar radial migration in gas-rich discs at high redshift

Abstract

Radial migration and dynamical heating redistribute stars within galactic discs and thereby modify the chemo-kinematic structure of their host galaxies. Usually, these secular processes are studied in N-body and hydrodynamical simulations of Milky Way analogues with stellar-dominated discs. In contrast, discs at high redshift are gas-rich, which may qualitatively change how secular evolution proceeds. We use the Nexus framework to construct and evolve a suite of isolated galaxies with fixed halo and disc mass but varying initial disc gas fraction, from 0% to 100%. We show that in gas-rich models, the root-mean-square change in stellar angular momentum is up to a factor of two larger than in gas-poor analogues and is accompanied by stronger radial and vertical heating, leading to enhanced radial mixing. We further dissect the role of gas in specific migration channels. For bar-driven migration, corotation resonance dragging dominates in gas-poor discs, whereas in gas-rich discs, stars more readily reach and accumulate at the outer Lindblad resonance, which acts as a barrier. The high radial mixing efficiency in gas-rich phases can flatten the stellar metallicity gradient relative to that of the initial gaseous disc within only a few orbital timescales. Together, these results imply that radial mixing in early, gas-rich discs is substantially more vigorous than in late-time, gas-poor discs, naturally producing distinct evolutionary tracks for chemically bimodal discs such as that of the Milky Way.

Figures (9)

• Figure 1: Top sets: the surface density maps of the gaseous component in the snapshots of galaxies at 1.2 Gyr and 2.0 Gyr (final snapshot) with turbulent gas in Set 1 of $f_{\rm gas}=20\%, 40\%, 60\%, 80\%$ from left to right. Bottom sets: same as the top, but they show the surface density maps of the pre-assembled disc stars in the final snapshots. All the panels are 24 kpc across.
• Figure 2: The root-mean-square variation of stellar orbital angular momentum normalised by the circular velocity of the galaxies. Top: the mean RMS variation of angular momentum as a function of the guiding radii of the stars in the final snapshot of the galaxies after $\Delta T \approx 1.5$ Gyr since the initial ($t =500~\rm Myr$) snapshot. Each solid line is coloured according to the gas fraction of the galaxy, and the yellow dashed line is for the analogue galaxies with isothermal gas of $f_{\rm gas}=20\%$. The vertical dash-dotted line labels the current corotation radius of the galaxy with $f_{\rm gas}=20\%$. Bottom: the time evolution of the angular momentum variation traced in different snapshots since time $\Delta T$ after the $500~\rm Myr$ snapshot, averaging among stars with guiding radius between $4-13$ kpc. The zoomed-in panel shows the region in the log--log scale, and the blue dotted line is the $\delta L_z\propto\Delta T^{1/2}$ line to indicate the power-law nature of the angular momentum variation in these galaxy models. A different choice of the guiding radius range does not qualitatively affect the trends.
• Figure 3: The mean relative density perturbation experienced by the stars over the $\sim1.5$ Gyr orbital times in the galaxies with different gas fractions, $f_{gas}$. Top: the mean total density perturbations averaged along the stellar orbits for stars in different guiding radii in the final snapshot. Each line is coloured according to the gas fraction of the galaxy. Bottom: the ratio of the relative density perturbation contributed by the gas and stars. The grey dashed line is the $\rm ratio=1$ line, for which regions above the line means that the gas component dominates the density perturbations, and regions below the line means that stars dominate the perturbations.
• Figure 4: Similar to Fig. \ref{['fig:RMS_Lz']} but shows the radial and time dependence of the radial and vertical heating-to-migration ratio of galaxies with different gas fractions. The heating-to-migration ratios, $\mathcal{H}$, are quantified as the ratio of the RMS variation of radial and vertical actions to the RMS variation of the azimuthal actions, $\mathcal{H}_{R,z} =\delta J_{R,z}/\delta L_z$, which the radial heating-to-migration ratio is shown on the left, and the vertical one is on the right. Top: the mean heating-to-migration ratios of stars as a function of the guiding radii in the final snapshot after $\sim1.5$ Gyr since the beginning. The vertical dashed lines and the bands are the current OLR radius of the galaxy with $f_{\rm gas}=20\%, 40\%$ and its $1\sigma$ region, which roughly corresponds to the radius for which the radial trends of $\delta J_{R,z}/\delta L_z$ change. Bottom: the time dependence of the heating-to-migration ratios, which we averaged over the stars between $4<R_g/\rm kpc<13$ in each snapshot. The grey shaded regions on the left denote the time that $\Delta T$ is less than an orbital period. The horizontal blue shaded regions denote the measured heating-to-migration ratio of the Milky Way (Frankel2020; Zhang et al. in prep.)
• Figure 5: The stellar surface density maps of simulation snapshots of the two galaxies in Set 2 at the bar formation moment and 2.4 Gyr after that. The galaxy without gas is shown on the left and the galaxy with $f_{\rm gas}=20\%$ is shown on the right. The red solid and dot-dashed circle denote the CR and OLR radius of the corresponding snapshots. The expansion of the CR and OLR radius from the bar formation to the final snapshot is due to the bar deceleration.