Cometary BAL Clouds?

If BAL clouds have optically thick cores surrounded by lower column density envelopes, then the envelopes will be more affected by radiation pressure than the cores and will be accelerated outwards from the quasar more than the core of the cloud will. This might form a tail analogous to a comet's dust tail, trailing the cloud opposite to its direction of motion. Such an effect might be detectable by looking to see if variable BAL troughs get stronger more quickly than they get weaker. That is, as a cometary BAL cloud moves across our line of sight, it should first produce a sharp jump in absorption from the core, followed by a slower decline from the larger envelope. That kind of variability was seen in X-rays by Maiolino et al. [that's where I got the idea] and in the UV at one point in FBQS 1408+3054 (see the 3rd-5th data points in Figure 3 of my paper); the SDSS DR5 and DR6 BAL catalogs should have many more examples to use to look for this effect (should also check the Gibson et al. 2008 and 2010 papers to see if they've already looked for this in their datasets).