Synthetic nebular emission from massive galaxies II: ultraviolet-line diagnostics of dominant ionizing sources

We compute synthetic optical and ultraviolet (UV) emission-line properties of galaxies in a full cosmological framework by coupling, in post-processing, new-generation nebular-emission models with high-resolution, cosmological zoom-in simulations of massive galaxies. Our self-consistent modelling accounts for nebular emission from young stars and accreting black holes (BHs). We investigate which optical- and UV-line diagnostic diagrams can best help to discern between the main ionizing sources, as traced by the ratio of BH accretion to star formation rates in model galaxies, over a wide range of redshifts. At low redshift, simulated star-forming galaxies, galaxies dominated by active galactic nuclei and composite galaxies are appropriately differentiated by standard selection criteria in the classical [OIII]$\lambda$5007/H$\beta$ versus [NII]$\lambda$6584/H$\alpha$ diagram. At redshifts $z \gt 1$, however, this optical diagram fails to discriminate between active and inactive galaxies at metallicities below $0.5\ Z_\odot$. To robustly classify the ionizing radiation of such metal-poor galaxies, which dominate in the early Universe, we confirm 3 previous, and propose 11 novel diagnostic diagrams based on equivalent widths and luminosity ratios of UV emission lines, such as EW(OIII]$\lambda$1663) versus OIII]$\lambda$1663/HeII$\lambda$1640, CIII]$\lambda$1908/HeII$\lambda$1640 versus OIII]$\lambda$1663/HeII$\lambda$1640, and CIV$\lambda$1550/CIII]$\lambda$1908 versus CIII]$\lambda$1908/CII$\lambda$2326. We formulate associated UV selection criteria and discuss some caveats of our results (e.g., uncertainties in the modelling of the HeII$\lambda$1640 line). These UV diagnostic diagrams are potentially important for the interpretation of high-quality spectra of very distant galaxies to be gathered by next-generation telescopes, such as the James Webb Space Telescope.
Comment: 23 pages, 10 figures, submitted to MNRAS