Model weights and training, validation, and test set images and masks for "Mapping one million small dams in Brazil reveals widespread environmental and policy impacts"

Annotated masks and Sentinel-1/-2 images split into training, validation, and test sets. Used for training convolutional neural network for small reservoir mapping. - manet_sentinel.ckpt: PyTorch model checkpoint file containing model weights. - annotations_locations_stats.csv: Basic information about annotation masks, including names, locations, and pixel-wise accuracy statistics of final model. Contains columns: name: File name, with .tif extension. Annotation file name is the same but with .png extension. set: train/val/test biome: Biome of center coordinates center_longitude, center_latitude: Coordinates of center of annotation xmin, xmax, ymin, ymax: Mask extents, in lat/lon coordinates true_positive_pixels, false_positive_pixels, true_negative_pixels, false_negative_pixels: Pixel-wise statistics based on final model predictions. - annotations.zip: Contains binary reservoir masks (0 is non-reservoir, 1 is reservoir) split into training, validation, and test sets, in png format. - images.zip: Contains Sentinel-1/-2 images in tif format split into training, validation, and test sets with the following bands: Blue Green Red Near-infrared Sentinel-1 SAR VV Sentinel-1 SAR VH NDVI NDWI Gao's NDWI MNDWI - manually_checked_detections.gpkg: We also include a shapefile of 5000 manually checked reservoir detections including their accuracy assessment and error type. Accuracy assessment ("label") is one of "True Positive", "Questionable", and "False Positive". Error type ("error_type") is one of: True Positive (i.e. not an error) Natural Water Body Non-Stream-Fed Impoundment (i.e. off-stream) Duplicate Detection Shadow Land Fragment of >50 ha reservoir