The analysis of the spatial distribution and the dynamics of green areas in the cities is of great importance today considering the increasing rate of urbanization and the need to ensure a comfortable and stable urban environment for the city dwellers. We tested methods of satellite image interpretation in GIS, including unsupervised and supervised classifications, while studying several Kharkiv urban parks and the city of Kharkiv as a whole. The use of automated machine learning methods allows to quickly identify land cover types and process large amounts of information, but the accuracy of the results may vary. The paper examines the changes in the area of greenery in three urban parks (a type of sports parks) in the city of Kharkiv after their reconstruction using the methods of unsupervised classification, such as Green City and Scene Classification Map on the EO Browser platform. In the period from 2017 to 2021, the reconstruction of parks resulted in a decrease in the area of green spaces. After the reconstruction, Saltivskyi urban park has less than 30% of the green area left. The paper also examines the possibility of the analysis of the spatial distribution of green areas in the city using unsupervised and supervised classifications. The methods used to classify satellite imagery can complement each other and provide a more encompassing picture that will help formulate recommendations to improve the spatial planning of green spaces in the study area.