Initially this work focusses on a review of the RING effect and its origin covering the last forty years. A large number of scattering processes have been assessed as potential contributors to the effect from which rotational RAMAN scattering was found to be the most important. The radiative transfer has been extended by this inelastical scattering process in order to determine its impact on the measured amount of light. Previous approaches were hampered by oversimplification. The presented theory of radiative transfer which accounts for rotational RAMAN scattering is complete and not restricted to a special scheme of solution.

In order to compensate spectral signatures induced by the RING effect for trace gas retrievals by means of so-called RING reference spectra, data of an extended radiative transfer model have been used. Besides the possibility of using model data semi-empirical RING reference spectra are utilized in trace gas retrievals. The interpretation of such spectra on the base of theoretical considerations shows weak points, which partially make the quality of these spectra questionable. Differences between modeled and measured RING reference spectra can therefore not only be traced back to deficiencies of the modeling of the effect.

In the course of the validation mostly good agreement of experimental- and model data with model data of this work has been found. Again it could be shown that rotational RAMAN scattering is the predominant source of the RING effect. Especially the comparison with data of the satellite-bourne sensor GOME (Global Ozone Monitoring Experiment ) pointed out good agreement. This is noteable, since the comparison was based on the optical depth and not on the RING reference spectra and therefore was free of critical assumptions.

Apart from describing the in-filling of solar FRAUNHOFER lines the RING effect fills-in also gasabsorption lines of appropriate width. The effect depends therefore not only on pressure, temperature and viewing-geometry, but also on the trace gas column itself. In order to account for this problem the separation of solar FRAUNHOFER and gasabsorption lines utilizing the HOTELLING transform has been performed. The method clearly shows that the first principle component corresponds to the in-filling of the solar FRAUNHOFER-, and the second principle component to the in-filling of gasabsorption lines. First results using both principle components in retrievals of NO₂ show good results.

The retrieval of O₃, NO₂ and BrO using different semi-empirical-, and modeled RING reference spectra lead to significant differences in the column-densities determined. Trials with modeled data resulted in errors of 2.5-5% for the vertical columns of O₃ and NO₂ if significantly different tracegas concentrations have been used for the optical depth and the modeled RING reference spectra. The neglect of a RING reference spectrum lead to errors of 6-10% for the vertical column (the evaluation of BrO is not possible without such reference spectrum). Therefore errors due to erroneous tracegas contents can be similar to completely neglecting the RING reference spectrum at all. The evaluation of BrO has been performed with experimental data only. Therefore the true values for the column density are not known. However, similarities to the NO₂ and O₃ model retrievals could be found, because differences of about 5% could be obtained using a RING reference spectrum with and without containing BrO absorption. Similar impact, as for the BrO evaluation is expected for OClO, HCHO and SO₂ retrieval since these gases have similar spectral structures. Furthermore they are retrieved utilizing their absorption bands in the UV, which is a spectral region that is strongly influenced by the RING effect.