Key points are not available for this paper at this time.
We present the results of a statistical investigation of broad emission-line profiles in 123 high-luminosity AGNs using high-quality data in the C IV λ1549 spectral region, originally obtained for absorption-line studies. Several systematic effects are observed: with increasing line width, the equivalent width and peak-to-continuum intensity ratio of C IV λ1549 decreases, the ratio of the equivalent width of the λ1400 feature to that of C IV λ1549 increases, the profile shape of C IV λ1549 changes from sharply peaked to more flat-topped, the line asymmetry changes, and there is an increase in the mean wavelength of the 1400 feature in the "rest frame" defined by the peak of C IV λ1549. These trends with line width can best be explained in terms of a range of relative strengths of two components making up each broad-line profile. The C IV λ1549 profile consists of a "core" with a characteristic width of ∼2000 km s-1 FWHM and an equivalent width that can differ significantly from object to object, plus a broad "base" component with a characteristic width of about 7000 km s-1, blueshifted by ∼1000 km s-1 relative to the core, and with a smaller range of equivalent widths. In this intermediate-width emission-line region (ILR) model the core emission arises from a region with a velocity dispersion intermediate between those of the narrow- and broad-line regions. Alternatively, we suggest a "bipolar" model consisting of a "disk" giving rise predominantly to low-ionization lines, plus a co-axial high-ionization outflow. This axisymmetric model incorporates viewing angle and projection effects to explain the main line-width correlations. The "ILR" model seems to provide a more natural explanation of the correlations presented here, and of the behavior of other lines reported in the literature. We compare the properties of the radio-loud and radio-quiet objects and find differences that we interpret as arising from a beamed optical-UV synchrotron continuum present in the radio-loud objects but absent in radio-quiet objects.
Wills et al. (Fri,) studied this question.