7.4. Dust and Reddening in the BLR
As discussed in the previous chapter, the agreement between models and observations of the BLR is very good for some lines and poor for others. For luminous AGNs, the comparison between the observations and the models (Fig. 12) suggests a smooth trend, in a sense that calculated ultraviolet line intensities, relative to L, are in better agreement with the observations compared with the calculated optical line intensities. The disagreement may thus be a continuos function of wavelength, which is typical of reddening by dust. A small amount of line-of-sight extinction (mV 0.6 mag.) can considerably reduce this discrepancy. Unfortunately, the situation is far from being simple. The uncertainty in current BLR models is more than enough to account for the above discrepancy and there are several other explanations for this discrepancy.
As for the low luminosity AGNs, there are strong indications that the faint, broad Balmer line wings observed in many of those, are heavily reddened. In view of the observational and theoretical limitations, the idea of reddening in BLRs must be investigated by looking for emission line ratios that are good reddening indicators.
Most broad emission lines are optically thick and their calculated intensity somewhat uncertain. In particular, the L / H / H ratio cannot be used as a reddening indicator because of the complicated line transfer and other uncertainties already discussed. The situation regarding the hydrogen Paschen lines is somewhat better, and the uncertainty involved in the calculated ratio of lines originating from a common upper level, such as P and H, is perhaps not as large. Out of all other emission lines, only two line pairs seem to be adequate reddening indicators. These are the HeII and the OI line pairs discussed below.
7.4.1 Reddening from HeII lines. As discussed in the previous chapters, the HeII spectrum of broad line AGNs is relatively simple and reliable calculations are already available. The complicated HeII L transfer does not influence the HeII Balmer and Paschen lines, and even the small optical depth in the HeII H line, at 1640Å, does not change the line ratios by too much. Thus, the theoretical Paschen and Balmer HeII lines can be compared with the observations to check for reddening. In particular, the theoretical calculations predict that
Currently, there are not enough reliable measurements of this line ratio to make any general statement about reddening in BLRs. There are some indications, in a few objects, that a small amount of reddening is indeed present. The observations are difficult because of the weakness of the HeII lines and the blending with nearby spectral features. They are also hampered by the large time variation in the intensity of the HeII lines in low luminosity AGNs.
The HeII10123 line has now been measured in several sources and can be used, given adequate infrared spectral resolution, as another reddening indicator for low redshift AGNs.
7.4.2 Reddening from OI lines. The calculated OI line spectrum is somewhat model dependent because of the fluorescence with L. However, most of the emission in the OI lines at 8446 and 1302Å (Fig. 6) is due to this process and there is only a little extra contribution to the OI1302 line due to collisional excitation. As a result, the line ratio is easy to calculate and it is a useful reddening indicator. Recent photoionization calculations suggest the following range for the theoretical line ratio:
Here again the measurements are very difficult to perform and there are only a handful of those. The little information available so far suggest some broad line reddening, by an amount which is highly uncertain.