Innovative land cover monitoring and change detection applications are being used by the Ministry of Housing, Territorial Planning and Environment of Uruguay (MVOTMA) with technical support from FAO for improved land and water management, sustainable agriculture and ecosystem services. Technical support initially focused on enhancing capacities in land cover monitoring, consistent series of land cover information being prerequisite for planning and management of natural resources. 

Online technical support was provided on integrating high resolution satellite data, implement an ML based classification algorithm and using cloud computing for strengthening existing land cover and crop monitoring capacities for the entire country. Several online trainings were provided on Land Cover Classification System, (LCCSv3), image processing (Optical and RADAR), object-based classification, machine learning and cloud computing (SEPAL and Google Earth Engine). 

Using two pilot areas and identifying the gaps and needs regarding available data, technical recommendations were provided for additional samples collection and using the cloud computing platform SEPAL platform as well as other platforms such as GEE (to extract Sentinel 1 statistics apart from the Sentinel 2 statistics), R based machine learning (ML) model to classify the land cover. The proposed methodological approaches used by national technical experts provided an overall accuracy of ~97%. The national level ML model will be enhanced to prepare the entire national land cover map as well as other remote sensing applications in support to sustainable land and water management. Other remote sensing applications were supported for crop (soja and maize) monitoring using parcel data, Sentinel 1 satellite images, GEE and Random Forest ML model. 

Technical collaboration between MVOTMA and FAO continues on developing remote sensing application for agricultural and land monitoring in support to sustainable land management and valorisation of ecosystem services.