Relevance of the work is caused by the necessity to create and to develop the techniques of mapping forest vegetation using satellite radar data. The main aim of the research is to study the possibility of classifying the coniferous and mixed forests of Baikal region based on polarimetric radar data and seasonal changes of polarimetric properties of vegetation in the decimeter range. Methods: transformation scattering matrix into a coherent and covariance matrix; polarimetric decomposition by the Claude-Pottier method; two- and three-parameter classification considering zonal distribution of the Claude-Pottier decomposition parameters; Freeman decomposition method to the volumetric, surface and double components of scattering; estimate of a biomass and tree density on the basis of the weights ratio of the components of Freeman decomposition; calculation of polarimetric features such as radar vegetation index, depolarization, etc.; comparative analysis of seasonal polarimetric characteristics. Results. The authors have studied seasonal changes of the polarimetric radar characteristics of the coniferous and mixed forests in Baikal region based on the autumn-winter data ALOS PALSAR in the decimeter range. Applying the Claude-Pottier and Freeman methods the polarimetric decompositions and classification based on the differences in a composition and a ratio of decomposition components were carried out. The authors revealed the considerable seasonal variations of scattering in vegetable and soil covers. It was ascertained that the dominant role of volume scattering in winter period decreases and surface scattering rises. All these indicate the corresponding variations of vegetative biomass. On the investigated site of pine forest the authors determined the decrease of the volume and double components of Freeman decomposition on average by 3 dB, which indicates a change in the conductivity properties of the forest medium at the negative temperatures.