We have studied the properties of Seyfert galaxies with high [OIII]5007 blueshifts ("blue outliers"), originally identified because of their strong deviation from the M BH − σ relation of normal, narrowline Seyfert 1 (NLS1) and broad-line Seyfert 1 (BLS1) galaxies. These blue outliers turn out to be important test-beds for models of the narrow-line region (NLR), for mechanisms of driving large-scale outflows, for links between NLS1 galaxies and radio galaxies, and for orientation-dependent NLS1 models. We report the detection of a strong correlation of line blueshift with ionization potential in each galaxy, including the measurement of coronal lines with radial velocities up to 500-1000 km s −1 , and we confirm a strong correlation between [OIII] blueshift and line width. All [OIII] blue outliers have narrow widths of their broad Balmer lines and high Eddington ratios. While the presence of nonshifted low-ionization lines signifies the presence of a classical outer quiescent NLR in blue outliers, we also report the absence of any second, non-blueshifted [OIII] component from a classical inner NLR. These results place tight constraints on NLR models. We favor a scenario in which the NLR clouds are entrained in a decelerating wind which explains the strong stratification and the absence of a zero-blueshift inner NLR of blue outliers. The origin of the wind remains speculative at this time (collimated radio plasma, thermal winds, radiatively accelerated clouds). It is perhaps linked to the high Eddington ratios of blue outliers. Similar, less powerful winds could be present in all Seyfert galaxies, but would generally only affect the coronal line region (CLR), or level off even before reaching the CLR. Similarities between blue outliers in NLS1 galaxies and (compact) radio sources are briefly discussed.

No abstract

Using a 100 ks XMM-Newton exposure of NGC 4051, we show that the time evolution of the ionization state of the X-ray absorbers in response to the rapid and highly variable X-ray continuum constrains all the main physical and geometrical properties of an AGN ''warm absorber'' wind. The absorber consists of two different ionization components, with a difference of %100 in ionization parameter and %5 in column density. By tracking the response in the opacity of the gas to changes in the ionizing continuum, we were able to constrain the electron density of the system. We find n e ¼ (5:8Y21:0) ; 10 6 cm À3 for the high-ionization absorber and n e > 8:1 ; 10 7 cm À3 for the low-ionization absorber. These densities require that the high-and low-ionization absorbing components of NGC 4051 must be compact, at distances 0.5Y1.0 lt-days (2200R S Y4400R S ) and <3.5 lt-days (<15;800R S ) from the continuum source, respectively. This rules out an origin in the dusty obscuring torus, as the dust sublimation radius is at least an order of magnitude larger ($12 lt-days). An accretion-disk origin for the warm absorber wind is strongly suggested, and an association with the high-ionization, He ii emitting, broad emission line region (radius <2 lt-days) is possible. The two detected phases are consistent with pressure equilibrium, which suggests that the absorber consists of a two-phase medium. A radial flow in a spherical geometry is unlikely, and a conical wind geometry is preferred. The implied mass outflow rate from this wind can be well constrained and is 2%Y5% of the mass accretion rate. If the mass outflow rate scaling with accretion rate is representative of all quasars, our results imply that warm absorbers in powerful quasars are unlikely to produce important evolutionary effects on their larger environment, unless we are observing the winds before they get fully accelerated. Only in such a scenario can AGN winds be important for cosmic feedback.

No abstract