*Astron. Astrophys. 359, 788-798 (2000)*
## 6. Conclusions and outlook
The expressions derived in this paper show that inside of an optically
very thick, differentially moving medium the total flux and the total
radiative acceleration in the limits of large and small velocity
gradients *w* are given on the one hand by simple functions of
*w* and on the other by expressions that depend on the plasma
properties only and that therefore can be precalculated without any
knowledge of the velocity field. For large *w*, essentially, a
wavelength average of the extinction coefficient enters. For the more
important case of small *w*, the free mean path is of primary
importance as in the static case, i.e. regions of low opacity
contribute most. This is in contrast to optically thin or partially
thick configurations such as e.g. stellar winds. The Doppler shifts
lead to derivatives, and since these
enter the flux and radiative acceleration in a highly non-linear way,
predictions even of the signs of the effects (as e.g. the change of
the flux with velocity) are uncertain without the knowledge of
specific details. We have therefore concentrated the analysis in this
paper on very simple spectral features and investigated a few more
complicated systems only numerically.
An alternative to this approach - which in addition provides a way
for *understanding* systems comprising many spectral lines - is
to represent the line positions, strengths, and shapes by a
*statistical* model and then solve the corresponding stochastic
equations. This will be done for a Poisson point process in a
subsequent paper.
© European Southern Observatory (ESO) 2000
Online publication: July 7, 2000
helpdesk.link@springer.de |