In this paper a review of correction methods and algorithms of atmospherically induced radiometric effects on satellite radar images has been presented. It is a first step in research approach leading to efficient elimination of “topographic effect” influencing the quality of automatic image classification. The paper arranges the models and approaches in correction procedures chains giving the theoretical base for further experimental works on different types of radar images. Generally, the correction methods rely on cosine models fulfilling Lambert’s law for visible range of solar spectra and for ideal, diffusely reflecting surface. The review of scientific papers on that matter leads to conclusions that theory is explaining only some aspects of natural surfaces backscattering mechanisms and that in many cases and approaches this topographic influence on radiometry is neglected by radar images users. For these reasons the next step of our research work will be testing and validating described algorithms on several radar datasets.