Extreme Dust Heating in Optically Star-Forming Galaxies

A complete census of supermassive black holes in the local universe is important, especially in low mass (log(stellar mass/solar masses) < 10) galaxies. It provides observational constraints on the black hole occupation fractions of low mass galaxies and broadens our understanding of the co-evolution of active galactic nuclei (AGN) and their host galaxies. Infrared selection criteria including [3.4]-[4.6] micron (W1-W2) color provides a useful method for detecting obscured AGN which may be missed in X-ray or optical surveys. Recent work has found that not only are there more AGN in low mass galaxies than would be predicted using optical selection criteria, but that the fraction of high W1-W2 (>0.5) galaxies is actually highest in the lowest mass galaxies. This could be evidence of a significant population of obscured AGN in low mass galaxies, but it is still unclear whether the dust heating that causes high W1-W2 color can only be caused by AGN or if stars alone are sufficient. This dissertation is a study of the demographics of high W1-W2 galaxies in the local universe and the AGN or star-forming nature of their nuclear activity.;First, the number density of z0.3, 0.5 and 0.8 are calculated as a function of r-band luminosity and stellar mass. Not only does the number density of high W1-W2 galaxies rise toward the lowest host mass regime in stark contrast to the mass distribution of optical AGN, but the red WISE population displays a bimodality in its luminosity and stellar mass functions. They are a combination of a high mass optical AGN and a low mass optically star-forming component. One optically normal, IR-red (W1-W2>1) galaxy (SDSS J1224+5555) was included in a pilot study of bulgeless, high W1-W2 galaxies which found that its X-ray flux is much lower than would be expected if it hosted an AGN. Decomposing its photometry with multiwavelength spectral energy distribution (SED) modeling revealed that it is impossible to reproduce the galaxy's mid-infrared fluxes without a substantial contribution from an obscured AGN. Lastly, the SED modeling strategy is extended to the low mass, W1-W2>0.8 galaxy sample. The ultraviolet through far-infrared SEDs of low mass, high W1-W2 galaxies can be explained by young stars with associated dust heating without the need to include an AGN. However the photometry of some optical AGNs in the test sample can also be fit with the same star and dust templates. While the emission of low mass, red W1-W2 galaxies is consisent with young stars and stellar dust heating, AGN activity cannot be conclusively ruled out. Deep and/or high energy X-ray observations or mid-infrared spectra of dwarf galaxies with high W1-W2 color are needed in order to determine whether these galaxies contain obscured AGN or their dust heating comes from stars alone.