We present the results of a variability study of ‘narrow-line Seyfert 1 galaxies’: in 10 objects out of 12 the optical line flux changed in one year, with no statistical differences in variability with respect to ‘normal’ Seyfert 1s. We discuss how models for NLSls are affected.

We present a brief review of emission-line velocity differences, and describe an ongoing project to determine the driving mechanisms responsible. We conclude with a brief outline of the use of velocity differences as probes of the conditions in the nuclear region of AGNs.

We consider in detail the two-phase BLR model with radial infall kinematics. Numerical calculations show that Lyα and C IVλ1549 profiles are blueshifted from their systemic rest-frame velocity, and their profiles are in qualitative agreement with the observations. It seems that the observed velocity differences between the high and low-ionization lines in the spectra of quasars can be understood in such a simple, physical, self-consistent model.

Many years of monitoring a sample of 10 AGNs with a median sampling rate of about one spectrum per week yields strong evidence that broad-line profile variations are not induced by reverberation effects, but rather signify real changes in the structure of the continuum-source and broad-line region complex, contrary to line flux variations, which do respond to continuum variations. If the profile variations indeed trace internal changes in the BLR, then the BLR cannot consist of the billions of small clouds as the standard model of the BLR prescribes. Rather, small-number statistics are necessary. The sample of AGNs also indicates there are three preferred ‘components’ in the line profiles. These can be explained as geometrical projection effects due to an anisotropic continuum irradiating an otherwise spherical BLR.

We study the C IV λ1549 emission-line profile variations in 18 Seyfert 1-type objects observed with the IUE satellite from 1978 to 1991. The core and the two wings of the line are integrated separately in order to study differences between the wings and the core and between the blue and the red wings. A principal component analysis is applied to spectra of each object. The results suggest the presence of a significant infall component in the line-emitting region of several AGN.

We postulate that all structure in broad lines can be explained by a central component (at the systemic redshift) and the addition of two ‘displaced components’, one blueshifted and the other redshifted. We have been able to successfully classify all Balmer-line profiles on this basis. 3C 390.3-type objects are merely examples where the shifts of the displaced components are unusually large. We believe that the displaced peaks are less prominent in the UV lines because the higher ionization lines are broader.

We have analyzed the HST FOS spectra of all quasars in the Stirpe (1990) high S/N line-profile sample and studied line-profile ratios as a function of radial velocity. Some quasars show no sign at all of NLR Lyα. We confirm that Hα is narrower than Lyα (after allowance for NLR contributions). The Lyα/Hα ratios in the cores of the broad lines are all close to or slightly less than case B and values predicted by single-cloud photoionization models. The Lyα/Hα ratio is surprisingly high in the blue wing. With only one exception, the ratios are equal to or greater than the case B value. Intrinsic reddening must be very small in most cases. We also briefly discuss other ratios.

We present preliminary results from a study of broad-line profiles in active galaxies. A simple model in which the emissivity is a broken power-law function of radius, and the BLR clouds emit anisotropically, yields very good fits to almost all the Ha profiles in our data base.

Optical monitoring data over several years in combination with the reconstructed 1-d response function for C IV obtained from the 1993 HST monitoring campaign of NGC 5548, reveal that: (a) radial motion does not dominate the gas kinematics (Korista et al. 1995); (b) line-profile variations occur in fixed regions of velocity space, a core and two wing components, are stochastic in nature, and appear to be unrelated to reverberation effects (Wanders & Peterson 1996). Wanders et al. (1995) showed that these variations are broadly consistent with a spherical BLR geometry populated with optically thick clouds following randomly inclined circular Keplerian orbits, and illuminated by an anisotropically emitting point source of ionizing continuum radiation. Here we provide a brief description of this model and summarize its basic properties. A more thorough analysis is presented in Goad & Wanders (1996).

Line-profile variations in the Seyfert galaxies NGC 5548 and NGC 4593 are discussed. The variations of individual line segments are different from line to line and from outburst to outburst.

We present new observations of the quasar 4C 37.43 in the course of a more general study of broad-line profiles in radio-loud AGN. During the most recent observations it became obvious that the broad profile Hβ had changed significantly. Difference spectra show that a new blue component has appeared on the Hβ profile. The centroid of this new component is displaced about 2500 km s−1 from the rest frame [O III] λ5007. A corresponding blueshifted He IIλ4686 component may have also appeared.

Emission-line profiles from both narrow eccentric rings and an extended disk around a black hole are presented. In both cases, a black-hole signature appears when the inclination of the emitting orbit is larger than 80°.

The bloated stars model for AGN broad-line emission proposes that this emission originates in the mass-loss envelopes or winds of giant stars. I present here a version of the model that simultaneously reproduces typical AGN broad emission-line ratios, line profiles, and linereverberation time lags without excessive stellar collisions. This can be achieved if (1) the bloated star envelope is about 10 times denser than that of normal supergiants, (2) the fraction of the bloated stars within the stellar population falls off as r−2, and (3) the size of the envelope is limited by the black-hole tidal forces.

By applying cross-correlation (CC) techniques to high-resolution (~ 6km s−1), high S/N (~ 400 at Hα line center) spectra of Mrk 335, we are able to put a lower limit of ~ 3 × 106 on the number of emitting clouds in this objects. This limit is applicable for clouds with T = 2 × 104 K and an optical depth of a few hundred in their Hα line. Current BEL models based on stellar atmospheres of bloated stars can be ruled out based on this lower limit.

We use thermal instability as a condensation mechanism to form the characteristic clouds of the BLR of quasars. We include in the total heat-loss function the influence of the heating caused by the damping of Alfvén waves. We study the damping mechanisms: nonlinear, resonance surface, and turbulent. From this analysis, we show that this heating mechanism plays an important role in the formation of these clouds.

A hydromagnetic wind model from a dusty molecular disk (Emmering, Blandford, & Shlosman 1992) was applied to the BLR in NGC 5548 and produced synthetic C IV line profiles as a function of time. The model C IV profiles are compared to data from the 1993 IUE/HST campaign and properties of the the BLR of NGC 5548 are inferred.

It is shown that, in the case of NGC 5548, photoionization models cannot account for line fluxes, line ratios, and line variations. The assumption that a fraction of the broad-line region is mechanically heated can solve the problem.

A thin disk illuminated by a central source will produce single-peaked broad emission lines if there is a wind emerging from the disk. The velocity gradient in the wind produces an anisotropic optical depth. For optically thick lines, the emission is strongest along directions perpendicular to the Keplerian velocity of the disk. The resulting line profiles are single peaked even though the emitting gas moves on essentially circular orbits. We argue that the broad emission lines seen in quasars, Seyferts, and luminous cataclysmic variables arise in disk winds.

We suggest that the broad-line regions (BLRs) of QSOs with broad emissionlines (BALs) are expanding, i.e., the clouds are undergoing radial outflow, as illustrated simply in Fig. 1. At least the following observational facts can be explained:

1. 1. Optical observations show that some moderate and high-redshift QSOs have BALs. The number of these QSOs is about 3–10% of all QSOs (Foltz et al. 1990). One can see in Fig. 2 that very low-redshift QSOs have no BALs.

2. 2. Almost all BAL QSOs exhibit zabs < zem.

3. 3. High-resolution observations reveal that the widths of the broad absorption lines are narrower than that of the corresponding emission line for all BAL QSOs, i.e. Δλ ab < Δλ em.

In 1980, You & Cheng (1980) suggested that in a dense gas medium, relativistic electrons will produce Čerenkov atomic or molecular emission lines of widths Δλ ≈ 1 – 10Å. The Čerenkov line is broader than a normal emission line and has small redshift (ΔZc ≡ Δλp/λlu ≈ 10–3), so the apparent velocity is about a few hundred km s−1. We refer to this as the ‘Čerenkov redshift’. In 1986, You & Cheng gave simplified formulae for Čerenkov line emission and the mechanism was confirmed by a series of elegant experiments (Xu et al. 1988, 1989). Recent progress in studies of AGNs, both in theory and observation, provide support for the Čerenkov line-emission model of the BLR of AGNs. In this paper, we prove that for a dense gas, if there are enough relativistic electrons, the Čerenkov line emission is strong enough to compare with the observations, and the Čerenkov line emission dominates over the spontaneous emission lines in the optically thick case.