Dry marginal agricultural land (DryMAL) potentially use as an alternative resource for crop production. DryMAL de-fined as land having low natural fertility due to its intrinsic properties and forming environmental factors. This study uses Sentinel-2A imagery to map the spatial extent, compare the result of the classification, and identify the change in DryMAL occupation. The area of study (461.9 km2) is part of Situbondo Regency and is located at the eastern part of East Java, In-donesia. Sentinel-2A image captured in dry-season of 2018 use for this study. Then, supervised image classification using a maximum likelihood algorithm use for image treatment and processing. Furthermore, 450 ground control points for train-ing areas collected during the field surveys. Five bands use in the classification process. The maps produced from the clas-sification process were then compared to the land-use map from the year 2000. The change in DryMAL occupation from 2000 to 2018 was calculated by comparing the classified and land-use map. Supervised classification yielded an overall accuracy of 95.8% and a kappa accuracy of 93.2%. The classification produced six (6) classes of land use: (1) forest, (2) pavement or built-up area, (3) irrigated paddy field, (4) non-irrigated rural area, (5) dry marginal land and (6) water body. Globally, during the last two decades, regional development led by the Regency occupied more DryMAL area for develop-ing plantation. The effort reduces the amount of non-irrigated and converting to the plantation, pavement areas, and irrigat-ed paddy-field.

Land use/land cover (LULC) maps are important datasets in various environmental projects. Our aim was to demonstrate how GEOBIA framework can be used for integrating different data sources and classification methods in context of LULC mapping.We presented multi-stage semi-automated GEOBIA classification workflow created for LULC mapping of Tuszyma Forestry Management area based on multi-source, multi-temporal and multi-resolution input data, such as 4 bands- aerial orthophoto, LiDAR-derived nDSM, Sentinel-2 multispectral satellite images and ancillary vector data. Various classification methods were applied, i.e. rule-based and Random Forest supervised classification. This approach allowed us to focus on classification of each class ‘individually’ by taking advantage from all useful information from various input data, expert knowledge, and advanced machine-learning tools. In the first step, twelve classes were assigned in two-steps rule-based classification approach either vector-based, ortho- and vector-based or orthoand Lidar-based. Then, supervised classification was performed with use of Random Forest algorithm. Three agriculture-related LULC classes with vegetation alternating conditions were assigned based on aerial orthophoto and Sentinel-2 information. For classification of 15 LULC classes we obtained 81.3% overall accuracy and kappa coefficient of 0.78. The visual evaluation and class coverage comparison showed that the generated LULC layer differs from the existing land cover maps especially in relative cover of agriculture-related classes. Generally, the created map can be considered as superior to the existing data in terms of the level of details and correspondence to actual environmental and vegetation conditions that can be observed in RS images.