Cyprus Terrace Detection and Classification Maps

This dataset contains the final raster outputs of the terrace detection and classification framework applied to the island of Cyprus. It includes maps representing terrace occurrence, terrace type, and terrace susceptibility derived from the machine-learning based object-based image analysis (OBIA). The dataset accompanies the submitted paper “Large-scale object-based detection of agricultural terraces: the example of Cyprus.” Terrace_binarypred_balthresh067.tif — Binary terrace occurrence map (terraced vs. non-terraced areas) obtained using a balanced threshold of 0.67 on model-predicted probabilities. Terrace_classpred_balthresh067.tif — Terrace type prediction map distinguishing broad agricultural, narrow agricultural, and reforestation terraces. Terrace_susceptibility_Cyprus.tif — Terrace susceptibility map showing the probability of terrace presence across the island. Format All files are GeoTIFF rasters in EPSG:32636 (WGS 84 / UTM zone 36N) coordinate reference system. Usage Please cite the associated paper and code repository when using this dataset.