Presentation #529.02 in the session “Quasars”.
Black hole-galaxy feedback is an established ingredient in cosmological structure formation simulations, but the contribution of quasars to feedback is not observationally determined. Among the candidates for quasar-mode feedback are broad absorption line (BAL) outflows and ionized emission-line outflows. These phenomena are found at similar distances from the central engine (parsecs to kiloparsecs), and they are both powered by photoionization of gas with similar physical conditions. It is conceivable that they are manifestations of the same outflowing gas, with the difference being that BAL quasar outflows are observed in the line of sight to the continuum source.
FeLoBAL quasars are a rare class of broad absorption line quasars (BALQs) that display absorption from MgII and FeII in their spectra. For 0.8<z<1.0, both [OIII] emission and FeII absorption are observed in a single SDSS spectrum. We present analysis of a sample of 30 low-redshift FeLoBAL quasars, many first identified using our convolutional neural net classifier FeLoNet. SimBAL analysis shows that the observed and predicted [OIII] emission may be related if the global covering fraction depends on distance from the central engine and Seyfert type.
We also analyzed a comparison sample of spectra from unabsorbed objects matched in 3-micron luminosity, signal-to-noise ratio, and redshift. Emission-line and principal-components analysis demonstrates that the two samples are not drawn from the same parent population. Half of the FeLoBAL sample are quasars with strong FeII and weak [OIII], with the relatively high L/LEdd traditionally expected in BALQ samples. But half of the FeLoBAL objects are Seyfert 1.5 and 1.8 objects (versus only 25% in the comparison sample). This latter group of FeLoBALQs is characterized by relatively low outflow velocities, luminosities, and Eddington ratios, all unexpected properties for FeLoBAL quasars. Thus, the range of accretion properties in FeLoBAL quasars is wider than expected.