Land Use and Land Cover Mapping using Multi-temporal Sentinel-1 Data and Machine Learning

Land use and land cover (LULC) maps are essential for effective policymaking and resource management. However, in many developing countries, including the Philippines, the lack of infrastructure, capacity, and technology poses significant challenges to generating reliable and timely LULC maps. With the availability of abundant Sentinel-1 data, various classification techniques have been explored to map LULC. This study focused on the proper preprocessing, feature extraction, and the use of RF50 for the classification of multi-temporal Sentinel-1A imagery. It also aimed to assess the performance of RF50 in LULC classification and mapping in the Philippine context. The methodology involved preprocessing the Sentinel-1 data, training and testing RF50 using a set of LULC classes, and developing an LULC map. The performance of RF50 is assessed using various metrics such as accuracy, precision, correlation and error rate.