There are  simulations of moving DACs in spectra of O stars.  We model different dencity distribution and investigate how the maximum of optical depth changes.
The regions of enhanced density are located near the surface of a star.

First model represents the situation when two "spots" are located near each other. The more dense spot is near \phi=0 with \rho=1.2, and less dense spot is on \phi=0.6 with \rho=0.8.  Thus the density ratio is  D=1.25.  As one may see from changes of optical depth with time the more dence stream has more high optical depth and with time, the picture keeps the same, i.e. more dense stream has higher optical depth.

The different colors refer to the different moments of time.
Black color t=0, blue t=10, green t=20, red t=50 and purple t=100 (in flow times).
Two features is due to two density enhancements. The spectral features behave just in the same way.

Let us now change location of these dense spots. Low density spot is on \phi=0 with \rho=0.8 and more dence spot is on \phi=0.6 with \rho=1.0, so D=1.25 as well . Surprisingly enought the maximum of optical depth does not belong all the time to the more dense component!!!. Colors are just the same as on the previous plots. The spectra reflect this situation! Thus, we see, that minimum of the flux does not bound with more dence stream! So it accelerate in a differnt way then spectral features themself. Therefore, it is obvious that the trace of minimum on the gray-scale diagrams of DACs, do not really reflect the motion of parcels of material. This diagnostic of DACs should be perform in a much more sensible way!

To proceed, let us trace the ratio of  minima of the fluxes of two DACs produced by two regions of enhnced dencsty with D=5 with time. The plot represents the
changes of    Min (F_1)/ Min (F_2) .

F_1 is an absorbtion componet produced by less dense stream and F_2 is an absorbtion component produced by more dense stream.  When F_1/F_2 < 1 it means 
that the DAC produced more dense stream is stronger. When F_1/F_2 > 1 it means  that the DAC originted because of the less dense component is more deep.
The simulations were performed  for different values \alpha for \alpha law for the velocity and \gamma -- rotational speed. As one may see    for all sets of parameters
the less dense component become dominant with time.

Next picture represents the nonuniform density distribution for ONE spot-stream.  The plot shows the time evolution of ONE DAC. One may clearly see that minimum of emission is drifting along the absorbtion line profile. Further, it looks like the minumum is accelerating more slowly than the bulk of absorbtion itself.  Nevertheless, when
density ratio is big enough, the secondary component is always much more shallow.

Next, we would explore the trace of the non symmetrical density distribution feature compare with symetrical. The density profile is shown here. The next plot represents
the absorption features as compare for symetricall and not symetrical density deistribution. The pink color covers the area restricted by a curve of minima absorbtion for non-symmetrical density distribution represented by this profile. The absorbtion features shown by blue and red color are due to a stream with symmetrical density distribution and all other parameters are the same. As it could be seen the area coverd by symmetruical features is excatly the same as pink non-symmetrical one.
Therefore, for this simulation we may conclude that non-symmetrical density distribution do not have crusial fluence on the acceleration of absorbtion components.
The parameters used for this simulatios are:
\gamma= 0.3 and \alpha=2;