Land use planning in the context of climate change, post-war reconstruction and adherence to European Union standards depends on reliable methods for predicting land use change. Using a significant amount of multi-temporal satellite data collected over a 7-year period from 2017 to 2023, we investigated the changes in land classes from one spatial and temporal transition state to the next one and to a future land use model. The MOLUSCE plug-in in QGIS used the following corrective (contextual) factors: digital elevation model (DEM), slope map, map of proximity to roads. Neural networks were used to calculate the forecast model. During the previous seven years, the area of different land uses had insignificant fluctuations, which indicates the relative stability of the territory. The classification data was verified using the Forest Plantation Plans of the educational and research forestry and information from the Public Cadastral Map. In the future, climate change may lead to a decrease in water levels, while the ability of ecosystems for self-regulation and the implementation of afforestation plans may increase the area covered by trees. For successful land management, urban planning and sustainable development, it is necessary to accurately predict trends in the structure of land. The lack of open sources of planned land use data (strategies, concepts, projects, etc.) makes it difficult to forecast the land use.