THE SPATIAL STRUCTURE OF MONO-ABUNDANCE SUB-POPULATIONS OF THE MILKY WAY DISK

The spatial, kinematic, and elemental-abundance structure of the MilkyWay's stellar disk is complex, and has been difficult to dissect with local spectroscopic or global photometric data. Here, we develop and apply a rigorous density modeling approach for Galactic spectroscopic surveys that enables investigation of the global spatial structure of stellar sub-populations in narrow bins of [alpha/Fe] and [Fe/H], using 23,767 G-type dwarfs from SDSS/SEGUE, which effectively sample 5 kpc < R-GC < 12 kpc and 0.3 kpc less than or similar to vertical bar Z vertical bar less than or similar to 3 kpc. We fit models for the number density of each such ([alpha/Fe] and [Fe/H]) mono-abundance component, properly accounting for the complex spectroscopic SEGUE sampling of the underlying stellar population, as well as for the metallicity and color distributions of the samples. We find that each mono-abundance sub-population has a simple spatial structure that can be described by a single exponential in both the vertical and radial directions, with continuously increasing scale heights (approximate to 200 pc to 1 kpc) and decreasing scale lengths (>4.5 kpc to 2 kpc) for increasingly older sub-populations, as indicated by their lower metallicities and [alpha/Fe] enhancements. That the abundance-selected sub-components with the largest scale heights have the shortest scale lengths is in sharp contrast with purely geometric thick-thin disk decompositions. To the extent that [alpha/Fe] is an adequate proxy for age, our results directly show that older disk sub-populations are more centrally concentrated, which implies inside-out formation of galactic disks. The fact that the largest scale-height sub-components are most centrally concentrated in the Milky Way is an almost inevitable consequence of explaining the vertical structure of the disk through internal evolution. Whether the simple spatial structure of the mono-abundance sub-components and the striking correlations between age, scale length, and scale height can be plausibly explained by satellite accretion or other external heating remains to be seen.