HI terminal velocity curves -- Lessons learned from N-body/hydrodynamic `surrogate' models of the Milky Way

Abstract

The development of an N-body/hydrodynamic `surrogate' model of the Milky Way (MW) - a model that resembles the MW in several key aspects after many Gyrs of evolution - would be extremely beneficial for Galactic Archaeology. Here we present four new `surrogate' models, all built with the Nexus framework. The simulations contain stars, dark matter and gas. Our most sophisticated model allows gas to evolve thermodynamically, and includes star formation, metal production, and stellar feedback. The other three models in this work have an isothermal gas disc. We examine these new simulations in the context of cold gas observations of the Galaxy. Our focus is the so-called `HI terminal velocity curve' - a heliocentric measurement of the maximum Vlos as a function of Galactic longitude, which dates back to the early days of radio astronomy. It is a powerful approach to indirectly estimating the gas dynamics because it does not require knowledge about the distance to individual gas clouds, which is difficult to estimate. A comparison of the terminal velocities and recovered rotation curve values in the simulations against observations suggests that our models are in need of further refinement. The gravitational torques associated with our synthetic bars are too strong, driving excessive streaming motion in the inner gas disc. This causes the simulated terminal velocity curves in the Galactic Quadrant I and IV to deviate substantially from each other, unlike what is seen in observed HI terminal velocities of the MW. We suggest possible ways forward for future models.

Figures (29)

• Figure 1: [id = anon]The left panel displays gas moving at $\mathrm{230 \unit{km/s}}$ along circular rings of different radii in $(x,y)$ space. The sun is located at $R_{\rm 0} = 8.2$ kpc and moves along the solar circle at speed $\Theta_0 = 240$ km s$^{-1}$. The purple ring of gas is positioned at the same radius as the sun. The right panel shows these rings mapped into $(\ell, V_{\rm los})$ space. The rings within the solar circle clearly appear as short diagonal lines confined to a narrow longitude range in the $(\ell, V_{\rm los})$ diagram. Rings outside the solar circle in $(x,y)$ appear as larger sin shaped curves in $(\ell, V_{\rm los})$. [id = anon]The curves in $(\ell, V_{\rm los})$ are overlaid on an observed distribution of H i emission in the Milky Way, adapted from reid2019trigonometric. The background greyscale shows the H i brightness temperature, integrated over the central 10 degrees of latitude.
• Figure 2: Left: Dashed circles indicate rings of radii $R = 1, 2, 3, 4, 5, 6$ and 7 kpc within the solar circle ($R_{\rm 0} = 8.2$ kpc; the location of the Sun is indicated by the black dot). For gas experiencing purely circular motion the maximum projected (i.e. terminal) velocity occurs at the point where the line-of-sight (red arrows) is a tangent to the relevant circular orbit. Right: The location of the gas with terminal velocities in $(x,y)$ space traces out an arc-like structure (red dots).
• Figure 3: Stellar surface densities for all the simulations evolving over time. We consider three simulation epochs: $t = 0, 2, 4$ Gyr. The black dashed line is identical in all panels, and it represents an exponentially declining surface density profile with $\Sigma_{\rm 0}\exp[-R/R_{d}]$ where $\Sigma_{\rm 0} = 10^{9} \unit{M_{\odot} \; kpc^{-2}}$ and $R_{\rm d} = 2.6 \unit{kpc}$. The black crosses are identical across panels, and they indicate the stellar surface density estimates from bovy2013direct, assuming an equal contribution of baryons and dark matter to local acceleration. Clearly, the stellar surface density profile in the simulations is broadly consistent with the observed stellar density profile.
• Figure 4: This shows face-on and edge-on stellar density projections of $t = 2.0$ Gyr from Model 1. We can see a distinct bar at the centre of the disc. The bar has already buckled and hence we also observe a boxy-peanut bulge when looking at the edge-on profiles.
• Figure 5: The first panel shows the distribution of gas within the solar circle for $t = 2.0$ Gyr. The stellar bar has been positioned at a $25 \degree$ inclination angle to the line connecting the Sun's position at $(-8.5,0)$ kpc and the galaxy's barycentre (at the origin). The black dotted line indicates the bar's semi-major axis. The dashed red and blue lines mark out the locations of the $18 \degree < \ell < 67 \degree$, and $-67 \degree < \ell < -18 \degree$ regions for Quadrant I and IV, respectively. These are the longitude and radius ranges for which the Quadrant I and IV H i terminal velocities of mcclure2016milky have been measured. The second to fourth panels present a zoomed-in view of the stellar density maps from Fig. \ref{['fig:stellar_density']}. The bar and its B/P structure are apparent.