Browsing by Subject "Star formation"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    An analysis of star formation in M31 using resolved stars and ultraviolet flux
    (2014-10) Simones, Jacob Edward
    We have used optical observations of resolved stars from the Panchromatic Hubble Andromeda Treasury (PHAT) to measure the recent (< 500 Myr) star formation histories (SFHs) of 33 FUV-bright regions in M31. The region areas ranged from ~ 104 to 106 pc2, which allowed us to test the reliability of FUV flux as a tracer of recent star formation on sub-kpc scales. The star formation rates (SFRs) derived from the extinction-corrected observed FUV fluxes were, on average, consistent with the 100-Myr mean SFRs of the SFHs to within the 1 σ scatter. Overall, the scatter was larger than the uncertainties in the SFRs and particularly evident among the smallest regions. The scatter was consistent with an even combination of discrete sampling of the initial mass function and high variability in the SFHs. This result demonstrates the importance of satisfying both the full-IMF and the constant-SFR assumptions for obtaining precise SFR estimates from FUV flux. Assuming a robust FUV extinction correction, we estimate that a factor of 2.5 uncertainty can be expected in FUV-based SFRs for regions smaller than 105 pc2, or a few hundred pc. We also examined ages and masses derived from UV flux under the common assumption that the regions are simple stellar populations (SSPs). The SFHs showed that most of the regions are not SSPs, and the age and mass estimates were correspondingly discrepant from the SFHs. For those regions with SSP-like SFHs, we found mean discrepancies of 10 Myr in age and a factor of 3 to 4 in mass. It was not possible to distinguish the SSP-like regions from the others based on integrated FUV flux.Starting from SFHs derived from the full PHAT photometric dataset, we have used stellar population synthesis to create maps of synthetic far- and near-ultraviolet (FUV and NUV) flux at sub-kpc resolution for the northeast quadrant of M31. The synthetic maps reproduced all of the main morphological features found in corresponding maps of observed FUV and NUV flux, including rings and large star-forming complexes. Comparing the flux maps pixel-by-pixel, we found the median synthetic-to-observed flux ratios to be 1.02 +0.74/-0.43 in FUV and 0.79 +0.35/-0.24 in NUV. The synthetic fluxes were therefore consistent overall with the observed fluxes in both filters. We used the observed fluxes and standard flux calibrations to derive star formation rate (SFR) maps, which we compared with a map of the mean SFRs over the last 100 Myr of the star formation histories (SFHs). We determined a lower limit of SFR ~ 10-5 Msun yr-1 below which the commonly assumed linear relationship between UV flux and SFR appears to break down. Above this limit, we found the median ratios of the flux-based SFRs to the mean SFRs to be 0.57 +0.47/-0.26 in FUV and 1.24 +0.88/-0.52 in NUV. Both the FUV and NUV flux-based SFRs were therefore consistent overall with the mean SFRs derived from the SFHs. Integrating over the entire mean SFR map, we found a global SFR of 0.3 Msun yr-1. The corresponding measurements from the flux-based SFR maps were factors of 0.74 (FUV) and 1.45 (NUV) of the global mean SFR value. It is not yet understood why the SFR ratios in the global case are larger than the median pixel-wise ratios. The primary source of uncertainty in both the synthetic flux maps and the flux-based SFR maps was most likely incomplete IMF sampling due to the small pixel areas. With the exception of the faintest areas of the galaxy, we did not identify any trends for flux or SFR with environment.