RESOLVE and ECO: Finding Low-metallicity z 0 Dwarf AGN Candidates Using Optimized Emission-line Diagnostics
Moffett, Amanda J.; Kannappan, Sheila J.; Norris, Mark A.; Polimera, Mugdha S.; Richardson, Chris T.; Bittner, Ashley S.; Ferguson, Carlynn; Eckert, Kathleen D.; Bellovary, Jillian M.
United States, United Kingdom
Abstract
Existing star-forming vs. active galactic nucleus (AGN) classification schemes using optical emission-line diagnostics mostly fail for low-metallicity and/or highly star-forming galaxies, missing AGN in typical z ~ 0 dwarfs. To recover AGN in dwarfs with strong emission lines (SELs), we present a classification scheme optimizing the use of existing optical diagnostics. We use Sloan Digital Sky Survey emission-line catalogs overlapping the volume- and mass-limited REsolved Spectroscopy Of a Local VolumE (RESOLVE) and Environmental COntex (ECO) surveys to determine the AGN percentage in SEL dwarfs. Our photoionization grids show that the [O III]/Hβ versus [S II]/Hα diagram (S II plot) and [O III]/Hβ versus [O I]/Hα diagram (O I plot) are less metallicity sensitive and more successful in identifying dwarf AGN than the popular [O III]/Hβ versus [N II]/Hα diagnostic (N II plot or "BPT diagram"). We identify a new category of "star-forming AGN" (SF-AGN) classified as star-forming by the N II plot but as AGN by the S II and/or O I plots. Including SF-AGN, we find the z ~ 0 AGN percentage in dwarfs with SELs to be ~3%-16%, far exceeding most previous optical estimates (~1%). The large range in our dwarf AGN percentage reflects differences in spectral fitting methodologies between catalogs. The highly complete nature of RESOLVE and ECO allows us to normalize strong emission-line galaxy statistics to the full galaxy population, reducing the dwarf AGN percentage to ~0.6%-3.0%. The newly identified SF-AGN are mostly gas-rich dwarfs with halo mass <1011.5 M ⊙, where highly efficient cosmic gas accretion is expected. Almost all SF-AGN also have low metallicities (Z ≲ 0.4 Z ⊙), demonstrating the advantage of our method.