Stars Born in the Wind II: Widespread Extra-planar Star Formation in M82's Halo

Abstract

Galaxies evolve in tandem with their environments -- mergers and gas inflows drive galaxy growth while galactic outflows launched by supernovae may seed the galactic environment with gas, metals, and energy, fueling star-formation far from the main bodies of galaxies. The formation histories of young stars in the stellar halos of nearby galaxies can help understand this interplay. We thus present the most detailed map to date of young stars in the stellar halo of M82, a starburst galaxy in the M81 Group that hosts a prototypical outflow, using Hubble Space Telescope (HST) and Subaru Hyper-Suprime Cam observations. We find widespread extraplanar populations of stars with ages $\lesssim630$ Myr, with clear detections of stars up to $\sim5$ kpc to the south in unique arc-like stellar features (Southern Arcs) and in a new stellar trail up to $\sim20$ kpc to the east (M82's Tail), originating from the Southern Arcs. We estimate a total halo star formation of $\sim4\times10^6\,M_\odot$ in the last $630$ Myr. Overall, the star formation history (SFH) of the M82 Tail is correlated with periods of heightened star cluster formation in the M82 disk, which suggests the influence of the starburst outflow. Further, the fraction of young stars decreases as we move away from M82 to the east. We forward a picture where the M82 Tail formed from ram pressure stripped gas arising from M82's westward motion, triggered by shocks from the outflow.

Figures (10)

• Figure 1: (Left) A Subaru HSC panoramic view of the M81 Group in resolved stars. Black points represent RGB stars and blue points represent young MS and BHeB stars. (Top Right) Zooming into a 40 kpc box around M82, the Southern Arcs and the M82 Tail consisting of $<400\,$Myr stars targeted in this work can be seen extending eastwards from M82. (Bottom Right) CMD of sources classified as stars in the Subaru HSC photometry. PARSEC Bressan2012-PARSEC isochrones (shifted assuming $(m-M)_0 = 27.8$) are over-plotted to broadly indicate the regions occupied by the $<400\,$Myr MS and BHeB stars (gray dashed line) and the ancient, RGB stars (dashed green line). These RGB stars are represented by $10\,$Gyr isochrones ([M/H] = 0.0, -1.0, -2.0).
• Figure 2: (Top) An HST resolved star view of M82 and it's halo with stars from the ANGST dataset Dalcanton2009, the Southern Arcs field Rao2025, and this work. The points in cyan represent populations of stars $\lesssim400\,$Myr, selected by excluding the RGB stars from the CMD of each dataset (dashed green regions in the CMDs below) with depths matched (F814W $<26$). The points in salmon represent the MS and BHeB stars from Subaru HSC. The M82 disk has been masked out to accentuate structures that follow the breakout of the outflow from the disk. The pink dashed line represents the approximate crowding limit for the Subaru dataset. (Bottom) CMDs of 'GST' sources from the Deep Halo Fields (including the Southern Arcs) arranged to reflect their relative positions on the sky. PARSEC Bressan2012-PARSEC isochrones with foreground extinction (shifted assuming $A_V= 0.2$ and $(m-M)_0 = 27.8$) are over-plotted to broadly explain the multiple stellar populations in each field. The $10\,$Gyr isochrones ([M/H] = 0.0, -1.0, -2.0) correspond to the old, RGB halo stars. For the $\leq630\,$Myr isochrones, we assume an [M/H]= -1.0.
• Figure 3: For all the Deep Halo Fields, (Left) Observed Hess diagram used by MATCH that balances number statistics on the CMD with time resolution in the SFH; (Center) best-fit model Hess diagram using PADUA stellar evolution models with complete AGB tracks Marigo2008Girardi2010; (Right) residual significance Hess diagram. The color bars in the left and center panels represent the number of stars per CMD bin. In the right panel, the scaling reflects the significance of each pixel in the residual relative to the standard deviation of a Poisson distribution. The RGB mask used is shown in gray in all panels along with the 50% completeness limit for each field.
• Figure 4: Best fit SFHs for each of the Deep Halo Fields using PADUA, MIST, and BASTI isochrones. Error envelopes represent the 68% confidence interval of random uncertainties. (Left) Absolute star formation rate (SFR) as a function of lookback time. (Right) Cumulative fraction of stars formed as a function of lookback time. The total stellar mass formed in the last $\sim630\,$Myr according to the PADUA models is also shown.
• Figure 5: A map $\lesssim400\,$Myr stars in the ANGST footprint (cyan) with cross-matches in the infrared from the Cibola (NIRCam) survey (salmon). The central disk has been masked out due to crowding and poor photometry. (Inset) CMD of ANGST GST sources in the outer disk and stellar halo with PARSEC isochrones overplotted. The dashed-gray region indicates the CMD mask used to exclude blends. Matched $\lesssim400\,$Myr stars are marked in salmon, and isochrones of various ages are overplotted.