Predicting Land Cover Map Changes in the Philippines for use in LULC-based Carbon Capture Monitoring using Deep Learning

In the area of Carbon Capture and Storage (CSS), monitoring plays an important role not only to determine if there is any anomalies in the release of CO2 in the atmosphere, but also to prepare for disasters and to plan better future developments in the industrial sector, transportation sector, real estate development, and other sectors. One way to monitor changes in the carbon cycle is by looking at Land Use and Land Cover (LULC) changes, since the primary methods of carbon capture and storage is by biological and geological sequestration. In this study, we designed a Deep Learning model that can predict land cover changes in the Philippine Land Cover Maps generated by the National Mapping and Resource Information Authority (NAMRIA). We evaluated our results and our model yielded a 78.64% overall accuracy and a Kappa coefficient of 0.725.