VALIDATION OF LAND DEGRADATION CARDS ON THE BASIS OF GEOSPATIAL DATA

Today there is a lot of satellite data and products based on it in the public domain. By integrating them with heterogeneous socio-economic information soil maps, model biophysical using modern machine learning methods approaches to geospatial processing, becomes possible create maps land degradation. Considering that classification productivity deforestation are main intellectual components degradation map, these affect overall reliability results. For their validation, necessary quality metrics determined work, corresponding calculations made. To evaluate cover independent test were used build confusion matrix, obtained areas crops compared statistical data. Agricultural was estimated time series Crop Growth Modeling System (CGMS) plant development, as well growth parameters development models. The LAI Map Accuracy Assessment (based CGMS) comparison Leaf Area Index (LAI) values modeled CGMS software framework ground measurement collected through surveys. Numerous experiments carried out assess models results an sample, which not at neural network training stage. Degradation for several years also analyzed validation respect productivity, particular region has undergone significant changes territory Ukraine.