The narrow line region of active galaxies - II. Relations between [O III] profile shape and other properties

Data from paper I have been combined with most previously published [O III] profile information to form a large sample of line width, asymmetry and kurtosis measurements for Seyfert galaxies. This sample has been used to investigate systematically the dependence of the [O III] profile shape on other properties. There are no correlations between [O III] line width and either the nuclear non-thermal luminosity, the relative strength of the Broad Line Region (BLR), or the ionization and physical state of the Narrow Line Region (NLR) gas. Previous results suggesting that Seyfert 2s have, on average, broader [O III] profiles than Seyfert 1s are shown to be a consequence of selection effects. For the Seyfert 1 sample, however, there are correlations between broad and narrow line width. This may indicate dynamical communication between the broad and narrow line regions although indirect effects may also be important. There are strong correlations between the [O III] line width and the NLR emission-line and radio luminosities (the latter first noted by Wilson & Willis). It is not yet clear which of these luminosities (emission-line or radio) is more fundamentally related to the velocity field. Real scatter in both correlations indicates that other variables are important, a conclusion which is reinforced by breakdown of the correlations at high luminosities. The Scyferts with linear kpc-scale radio morphology have steeper sided profiles (higher kurtosis) than those with diffuse or core–halo structure although the overall line widths and asymmetries are similar. Radio ‘jets’ or plasmoids may therefore perturb the NLR velocity field but probably do not dominate it. There is no strong tendency for edge-on galaxies to have broader lines. If an NLR disc is coaligncd with the galaxy then an upper limit of 50 per cent can be set for the contribution to the line width due to motion in that plane (e.g. rotation). The loose anticorrelation between excitation and profile asymmetry first noted by Hcckman et al. is confirmed for the present sample. However, no correlation is found between Balmer decrement and profile asymmetry. This may suggest a thick opacity source internal to the emitting clouds favouring an inflow model. Profile asymmetry and kurtosis are also unrelated to either non-thermal or emission-line luminosities or the relative strength and kinematics of the BLR. Combining these results with those from paper I, it is argued that motion in the bulge potential may dominate the NLR velocity field. The narrower lines found in Starburst galaxies would then suggest that NLR gas is not simply photoionized star formation material but may instead be a physically and kinematically distinct component.