About the Service
The CopPhil Land Cover Mapping Service provides solutions for a wide range of practical applications across various domains, including risk assessment and land use monitoring. Products developed for the service could support the identification and assessment of risks in urban and rural areas. Combined with auxiliary data such as Digital Elevation Models (DEM), they can be useful to detect vulnerable zones, such as flood-prone regions, landslide-risk terrains, or fire-susceptible areas. Decision-makers can then implement targeted mitigation strategies.
Monitoring changes in land use over time is critical for sustainable development and environmental conservation. Developed in close collaboration with PhilSA and NAMRIA, the service’s annual updates allow stakeholders to track urban expansion and its encroachment on natural habitats. This way, the land cover product provides a valuable tool for decision-makers to address environmental challenges.
Standalone Land Cover mapping
The Standalone Land Cover Mapping product generates a detailed map of land cover at a 10-metre resolution using the IOTA² classification toolbox. The mapping process draws on a combination of Sentinel-2 optical imagery, Sentinel-1 radar data, and elevation data, alongside a reference dataset used to train the classification model and a defined boundary of the area of interest.
To distinguish between land cover classes, the service analyzes how different surfaces, such as crops, forests, or urban areas, reflect light differently over time, drawing on up to 75 Sentinel-2 images per classification.
Land Cover Merge map
The Land Cover Merge product brings together the Standalone Land Cover, Forest Type, and Crop Type maps into a single, unified land cover map. For each location, a set of decision rules determines which classification source is most reliable, using the Standalone Land Cover map as the base and updating individual classes, such as perennial crops, where other sources provide greater confidence. The result is refined using minimum mapping unit parameters defined by NAMRIA, ensuring consistency with national standards.
Land Cover Change map
The Land Cover Change map identifies areas where land cover has changed over time, and classifies each detected change using the same land cover classes and minimum mapping unit parameters applied across the other CopPhil land cover products.
This product uses Sentinel-2 L2A and Sentinel-1 GRD images from the CopPhil mirror site to produce a time series covering a 3-year period. Changes are identified using the BFASTmonitor algorithm, classified through a machine-learning model, and refined by filtering out small or isolated detections to reduce noise in the final output.
