These datasets were generated from the research article "Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia". Here we publish the yearly time series maps showing land use following deforestation across Ethiopia produced at 10m resolution covering the year 2001 to 2015 (15 years). The land use maps are derived from sentinel-2 images using U-Net deep neural network architecture enhanced with attention. The dataset has eleven land use classes with values ranging from 1 -11, where 1: Larger-scale cropland, 2: Pasture, 3: Mining, 4: Small-scale cropland, 5: Roads, 6: Other land with tree cover, 7: Plantation forest, 8: Coffee, 9: Settlement, 10: Tea plantation, and 11: Water. The data is accompanied by the sld and qml file in a zip folder (Visualisation_layer_descriptor_sld_and_qml) to aid in visualisation or legend creation. Data Visualisation at 10m resolution: https://robertnag82.users.earthengine.app/view/deforestationdriverethiopia Zoom to this location (Longitude:35.27458, Latitude: 7.3291) to visualize expansion of coffee plantation, small and large scale croplands in a degraded forest. Zoom to this location (Longitude:35.48129, Latitude: 7.84093) to visualize expansion of Tea plantation after deforestation. File Naming: Landuse_dl_10m_s_20010101_20011230_af_epsg.4326_v20240201.tif Generic variable name: Landuse Method of data production: dl (Deep learning) Position in the probability distribution / variable type: c Spatial support in m: 10m Depth reference at surface ("s"): s Time reference begin time (YYYYMMDD): i.e. 20010101 Time reference end time: 20011230 Bounding box (2 letters max): af (Refering to Africa) EPSG code: epsg.3035 Version code i.e. creation date: v20240201 JSON field representing the legend. This is important for the visualization of the legend. [ { "Large-scale cropland": "1", "color": "#FFFF00" }, { "Pasture": "0", "color": "#808080" }, { "Mining": "1", "color": "#FFC0CB" }, { "Small-scale cropland": "120", "color": "#F39C12" }, { "Roads": "1", "color": "#800000" }, { "Other land with tree cover/Regrowth ": "0", "color": "#008000" }, { "Plantation forest": "1", "color": "#808000" }, { "Coffee": "120", "color": "#008080" }, { "Settlement": "1", "color": "#FF0000" }, { "Tea plantation": "0", "color": "#3CB371" }, { "Water": "1", "color": "#0000FF" } ]

These datasets were generated from the research article "Using high-resolution imagery and deep learning to classify land-use following deforestation: a case study in Ethiopia". Here we publish the yearly time series maps showing land use following deforestation across Ethiopia produced at 10m resolution covering the year 2001 to 2015 (15 years). The land use maps are derived from sentinel-2 images using U-Net deep neural network architecture enhanced with attention. The dataset has eleven land use classes with values ranging from 1 -11, where 1: Larger-scale cropland, 2: Pasture, 3: Mining, 4: Small-scale cropland, 5: Roads, 6: Other land with tree cover, 7: Plantation forest, 8: Coffee, 9: Settlement, 10: Tea plantation, and 11: Water. The data is accompanied by the sld and qml file in a zip folder (Visualisation_layer_descriptor_sld_and_qml) to aid in visualisation or legend creation. Data Visualisation at 10m resolution: https://robertnag82.users.earthengine.app/view/deforestationdriverethiopia