We are exploring a spectroscopic unification for all types of broad-line emitting AGNs. The four-dimensional Eigenvector 1 (4DE1) parameter space organizes quasar diversity in a sequence primarily governed by Eddington ratio. This paper considers the role of C iv k1549 measures as 4DE1 diagnostics. We use HST archival spectra for 130 sources with S/N high enough to permit reliable C iv k1549 broad-component measures. We find a C iv k1549 BC profile blueshift that is strongly concentrated among (largely radio-quiet [RQ]) sources with FWHM( H BC ) P 4000 km s À1 (which we call Population A). Narrow-line Seyfert 1 (NLSy1; with FWHM H 2000 km s À1 ) sources belong to this population but do not emerge as a distinct class. The systematic blueshift, widely interpreted as arising in a disk wind/outflow, is not observed in broader line AGNs (including most radio-loud [RL] sources), which we call Population B. We find new correlations involving FWHM(C iv k1549 BC ), C iv k1549 line shift, and equivalent width only among Population A sources. Sulentic et al. suggested C iv k1549 measures enhance an apparent dichotomy between sources with FWHM(H BC ) less and greater than 4000 km s À1 , suggesting that it has more significance in the context of broad-line region structure than the more commonly discussed RL versus RQ dichotomy. Black hole masses computed from FWHM C iv k1549 BC for about 80 AGNs indicate that the C iv k1549 width is a poor virial estimator. Comparison of mass estimates derived from H BC and C iv k1549 reveals that the latter show different and nonlinear offsets for Population A and B sources. A significant number of sources also show narrow-line C iv k1549 emission that must be removed before C iv k1549 BC measures can be made and interpreted effectively. We present a recipe for C iv k1549 narrow-component extraction.

No abstract

[LaTeX removed] Recent work has shown that it is possible to systematize quasar spectral diversity in a parameter space called ``Eigenvector 1'' (E1). We present median AGN spectra for fixed regions of the E1 (optical) parameter space (FWHM(H-beta) vs. equivalent width ratio RFE=W(FeII4570)/W(H-beta). Comparison of the median spectra for different regions show considerable differences. We suggest that an E1-driven approach to median/average spectra emphasizes significant differences between AGN, and offers more insights into AGN physics and dynamics than a single population median/average derived from a large and heterogeneous sample of sources. We find that the H-beta broad component line profile changes in shape along the E1 sequence both in average centroid shift and asymmetry. While objects with FWHM(H-beta)< 4000 km/s are well fitted by a Lorentz function, AGN with FWHM(H-beta)> 4000 km/s are well fitted if two broad line components are used: a broad (the "classical" broad line component) and a very broad/redshifted component.Comment: 1 table + 3 figures, accepted for publication in ApJ

We investigate the radial velocity difference between the [OIII]λλ4959,5007 and Hβ lines for a sample of ≈ 200 low redshift AGN. We identify seven objects showing an [OIII]λ5007 blueshift relative to Hβ with amplitude larger than 250 km s −1 (blue "outliers"). These line shifts are found in sources where the broad high ionization lines (e.g. CIVλ1549) also show a large systematic blueshift. Such blueshifts occur only in the population A region of the Eigenvector 1 parameter domain (that also contains NLSy1 sources). We suggest that [OIII]λλ4959,5007 blueshifts are also associated with the high ionization outflow originating in these sources. This is a direct kinematic linkage between narrow and broad line emitting gas.