3.2. Photoionization Models
Photoionization is the most likely source of excitation for the emission line gas in AGNs. The intense, nonstellar continuum is directly observed in many objects, in some cases well beyond the Lyman edge, and there is nothing to prevent this radiation from interacting with the surrounding material. Direct evidence for this is the correlated line and continuum variations in many Seyfert 1 galaxies, where the line luminosity changes are clearly a response to the continuum luminosity changes.
Some observed line ratios give further support to the photoionization assumption. For example, the CIII 977 / CIII] 1909 ratio is a good BLR temperature indicator for densities below about 1010 cm-3. The observed line ratio in several quasars (an upper limit usually), indicates an electron temperature below 25,000K in the C+2 zone. This is well below the temperature required to ionize carbon by collisions to C+2, and very typical of the temperature in a photoionized nebula.
An important concept in this kind of modeling is that of a "cloud", introduced to distinguish small individual entities in the emission line region. As explained earlier, the broad line gas is optically thick to the ionizing radiation, as deduced from the presence of strong MgII and FeII lines. On the other hand, in many objects the low and the high ionization lines have similar profiles, indicating that the line ratios are about the same throughout the emission line region. Thus a reasonable assumption is that there are many individual optically thick clouds, each one producing all the observed lines in roughly the same proportion. Later on we show that the number of clouds is very large and they occupy only a tiny part of the volume of the emission line region. An alternative picture that has been considered involves a spherical shell around the ionizing source, producing high excitation lines from its inner part and low excitation lines from its outer part. The total continuum optical depth of the shell can be large but it is hard to imagine a dynamical situation in which the high and the low excitation lines would have similar profiles. Accepting the cloud picture, we are led to the conclusion, based on the very smooth line profiles, that the number of clouds must be very large. There is no direct way to obtain information about the shape and dimension of individual clouds and some assumptions about it must be made when constructing the model.