PI: Victor P Debattista
We showed (Debattista et al. 2025, MNRAS, 537, 1620) that the azimuthal variations in mean stellar metallicity observed in the Milky Way can be attributed to spiral structure. We demonstrated that these azimuthal variations can be traced to more metal-rich stars being, on average, kinematically cooler. Since cooler populations support spiral structure more strongly, they tend to be more concentrated at the peak densities. While kinematically cool stars are typically younger, we showed that azimuthal metallicity variations are present in stars of all ages, not just the youngest ones. The mean metallicity variations are matched by similar variations in mean stellar age and alpha-abundance. Curiously however, if we measure the instantaneous radial action, JR, under the usual axisymmetric Stackel approximation, we find that low JR does not correspond to regions of high spiral density. Indeed variations in JR do not correlate very strongly with the variations in metallicity, age, local density, etc. This is counter to the expectation that the driver of the azimuthal variations is the different response of stars with different radial random motions to a perturbing spiral. However, if we average JR over a period long enough to ensure that stars will have drifted through the spiral structure, we find strong correlations between <JR> and the stellar population properties. Thus we caution that careful studies of the dynamics of spiral structure using both simulations and observations requires that the effects of the deviations from axisymmetry be taken into account when computing actions.
The attached figure shows the mean metallicity, <[Fe/H]> (colours) and density (contours) in our model. The <[Fe/H]> peaks and minima are highly correlated with the spiral peaks and troughs.
We (Deg et al. 2025, MNRAS 542, 464) developed a simulation campaign aimed at producing a more accurate realisation of the Milky Way. Previous attempts at building such a model have required the addition of a massive classical bulge to prevent the bar that forms from being too large. But the Milky Way very likely lacks any such bulge, instead having a bulge that formed purely via the secular evolution of the bar. In order to build our model, we assumed that the disc was more centrally concentrated than the usually assumed exponential. We chose a Sersic profile, because of its wide use in astronomy. The bars that formed in our simulations intially had a reasonable size. Nevertheless, fierce dynamical friction with the dark matter halo provoked runaway growth, leading to unrealistically large and slow bars. We were able to tame this growth by including a modest gas fraction. As a result of this addition, our fiducial model formed a bar with comparable size, angular pattern speed, and X-shaped structure as the Milky Way. We are now using these models to prepare for the installation of the MOONS instrument on the Very Large Telescope by predicting the kinematics of the bar.
