Recent spectroscopic observations in the Milky Way suggest that the chemically defined thick disk (stars with high [α/Fe] ratios and thus old) has a significantly smaller scale-length than the thin disk. This is in apparent contradiction with observations of external edge-on galaxies, where the thin and thick components have comparable scale-lengths. Moreover, while observed disks do not flare (scaleheight does not increase with radius), numerical simulations suggest that disk flaring is unavoidable, resulting from both environmental effects and secular evolution. Here we address these problems by studying two different suites of simulated galactic disks formed in the cosmological context. We show that the scale-heights of coeval populations always increase with radius. However, the total population can be decomposed morphologically into thin and thick disks, which do not flare. We relate this to the disk inside-out formation, where younger populations have increasingly larger scalelengths and flare at progressively larger radii. In this new picture, thick disks are composed of the imbedded flares of mono-age stellar populations. Assuming that disks form inside out, we predict that morphologically defined thick disks must show a decrease in age (or [α/Fe] ratios) with radius and that coeval populations should always flare. This also explains the observed inversion in the metallicity and [α/Fe] gradients for stars away from the disk midplane in the Milky Way. The results of this work are directly linked to, and can be seen as evidence of, inside-out disk growth.

We present an overview of the GHOSTS survey, the largest study to date of the resolved stellar populations in the outskirts of disk galaxies. The sample consists of 14 disk galaxies within 17 Mpc, whose outer disks and halos are imaged with the Hubble Space Telescope Advanced Camera for Surveys (ACS). In the first paper of this series, we describe the sample, explore the benefits of using resolved stellar populations, and discuss our ACS F606W and F814W photometry. We use artificial star tests to assess completeness and use overlapping regions to estimate photometric uncertainties. The median depth of the survey at 50% completeness is 2.7 mag below the tip of the red giant branch (TRGB). We comprehensively explore and parameterize contamination from unresolved background galaxies and foreground stars using archival fields of high-redshift ACS observations. Left uncorrected, these would account for 10 0.65×F814W−19.0 detections per mag per arcsec 2. We therefore identify several selection criteria that typically remove 95% of the contaminants. Even with these culls, background galaxies are a significant limitation to the surface brightness detection limit which, for this survey, is typically V ∼ 30 mag arcsec −2. The resulting photometric catalogs are publicly available and contain some 3.1 million stars across 76 ACS fields, predominantly of low extinction. The uniform magnitudes of TRGB stars in these fields enable galaxy distance estimates with 2%-7% accuracy.

We study the stellar halo colour properties of six nearby massive highly inclined disk galaxies using HST ACS and WFC3 observations in both F 606W and F 814W filters from the GHOSTS survey. The observed fields probe the stellar outskirts out to projected distances of ∼ 50 -70 kpc from their galactic centre along the minor axis. The 50% completeness levels of the colour magnitude diagrams are typically at two mag below the tip of the red giant branch. We find that all galaxies have extended stellar haloes out to ∼ 50 kpc and two out to ∼ 70 kpc. We determined the halo colour distribution and colour profile for each galaxy using the median colours of stars in the RGB. Within each galaxy we find variations in the median colours as a function of radius which likely indicates population variations, reflecting that their outskirts were built from several small accreted objects. We find that half of the galaxies (NGC 0891, NGC 4565, and NGC 7814) present a clear negative colour gradient in their haloes, reflecting a declining metallicity; the other have no significant colour or population gradient. In addition, notwithstanding the modest sample size of galaxies, there is no strong correlation between their halo colour/metallicity or gradient with galaxy's properties such as rotational velocity or stellar mass. The diversity in halo colour profiles observed in the GHOSTS galaxies qualitatively supports the predicted galaxy-to-galaxy scatter in halo stellar properties; a consequence of the stochasticity inherent in the assembling history of galaxies.

We analyzed the radial surface brightness profile of the spiral galaxy NGC 7793 using HST/ACS images from the GHOSTS survey and a new HST/WFC3 image across the disk break. We used the photometry of resolved stars to select distinct populations covering a wide range of stellar ages. We found breaks in the radial profiles of all stellar populations at 280 (∼5.1 kpc). Beyond this disk break, the profiles become steeper for younger populations. This same trend is seen in numerical simulations where the outer disk is formed almost entirely by radial migration. We also found that the older stars of NGC 7793 extend significantly farther than the underlying H i disk. They are thus unlikely to have formed entirely at their current radii, unless the gas disk was substantially larger in the past. These observations thus provide evidence for substantial stellar radial migration in late-type disks.

We study the properties of the stellar populations in M81's outermost part, which hereafter we will term the stellar halo, using Hubble Space Telescope (HST) Advanced Camera for Surveys observations of 19 fields from the GHOSTS survey. The observed fields probe the stellar halo out to a projected distance of ∼ 50 kpc from the galactic center. Each field was observed in both F606W and F814W filters. The 50% completeness levels of the color magnitude diagrams (CMDs) are typically at 2 mag below the tip of the red giant branch (TRGB). Fields at distances closer than 15 kpc show evidence of disk-dominated populations whereas fields at larger distances are mostly populated by halo stars. The RGB of the M81's halo CMDs is well matched with isochrones of ∼ 10 Gyr and metallicities [Fe/H] ∼ −1.2 dex, suggesting that the dominant stellar population of M81's halo has a similar age and metallicity. The halo of M81 is characterized by a color distribution of width ∼ 0.4 mag and an approximately constant median value of (F606W − F814W ) ∼ 1 mag measured using stars within the magnitude range 23.7 F814W 25.5. When considering only fields located at galactocentric radius R > 15 kpc, we detect no color gradient in the stellar halo of M81. We place a limit of 0.03 ± 0.11 mag difference between the median color of RGB M81 halo stars at ∼15 and at 50 kpc, corresponding to a metallicity difference of 0.08 ± 0.35 dex over that radial range for an assumed constant age of 10 Gyr. We compare these results with model predictions for the colors of stellar halos formed purely via accretion of satellite galaxies. When we analyze the cosmologically motivated models in the same way as the HST data, we find that they predict no color gradient for the stellar halos, in good agreement with the observations.