{"success":true,"data":{"author":"Daniele Rege Cambrin, Luca Colomba and Paolo Garza, Politecnico di Torino","bucketAccessType":"public","bucketName":"DarthReca/california_burned_areas","bucketPrefix":"","category":"Earth Observation & Geospatial","cloudProvider":"huggingface","createdAt":"2026-11-17T14:12:17.962Z","description":"CaBuAr (California Burned Areas) is a burned-area segmentation dataset of Sentinel-2 scenes acquired before and after wildfires in California, with binary ground-truth masks derived from the California Department of Forestry and Fire Protection (Cal Fire) fire perimeters mapped onto each scene. Full scenes are 5490 x 5490 pixels with all 12 spectral bands for the pre-fire and post-fire images; a pre-patched variant cuts them into 512 x 512 patches, keeping only patches with at least one burned pixel. Data are HDF5 files compressed with BZip2 through hdf5plugin (required to read them), grouped by fold and wildfire uid with pre_fire, post_fire and mask arrays, and provided in raw and normalized (0 to 255) versions under raw/ and normalized/; metadata.parquet gives coordinates, CRS and timestamps. There are 5 random folds (0 to 4), and the repo also includes the ChaBuD challenge test file. Load with h5py, the Hugging Face loading script (post-fire and pre-post-fire configurations) or TorchGeo's CaBuAr dataset class. Released under CDLA-Permissive-2.0 by the authors; cite Cambrin, Colomba and Garza, 'CaBuAr: California burned areas dataset for delineation', IEEE Geoscience and Remote Sensing Magazine 2023.","displayTitle":"CaBuAr: California Burned Areas, Pre- and Post-Fire Sentinel-2 Scenes with Cal Fire Masks","huggingFaceUrl":"https://huggingface.co/datasets/DarthReca/california_burned_areas","id":"376d6c75-ff8a-5117-b5b2-22874adc613e","license":"CDLA-Permissive-2.0","licenseEvidenceUrl":"https://huggingface.co/datasets/DarthReca/california_burned_areas","objectCountEstimate":45,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2401.11519","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"DarthReca/california_burned_areas\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"DarthReca/california_burned_areas\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=DarthReca, license:cdla-permissive-2.0, size_categories:n<1K, downloads=4848"],"format":"mixed","generatedAt":"2026-11-17T14:23:33.715Z","layout":{"example":"metadata.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"metadata.parquet","truncatedAt":512}],"sampleObjects":["california_burned_areas.py","metadata.parquet","normalized/complete/california_0.hdf5","normalized/complete/california_1.hdf5","normalized/complete/california_10.hdf5"],"sidecars":[{"bytes":"---\nlicense: cdla-permissive-2.0\ntask_categories:\n- image-segmentation\npretty_name: California Burned Areas\nsize_categories:\n- n<1K\ntags:\n- climate\n---\n# California Burned Areas Dataset\n\n**Working on adding more data**\n\n## Dataset Description\n\n- **Paper:** [Pre-Print](https://arxiv.org/abs/2401.11519) and [Version of Record](https://ieeexplore.ieee.org/document/10261881)\n\nYou can find an official implementation on [TorchGeo](https://torchgeo.readthedocs.io/en/stable/api/datasets.html#non-geospatial-datasets).\n\n### Dataset Summary\n\nThis dataset contains images from Sentinel-2 satellites taken before and after a wildfire. \nThe ground truth masks are provided by the California Department of Forestry and Fire Protection and they are mapped on the images.\n\n### Supported Tasks\n\nThe dataset is designed to do binary semantic segmentation of burned vs unburned areas.\n\n## Dataset Structure\n\nWe opted to use HDF5 to grant better portability and lower file size than GeoTIFF.\n\n*WIP*: You can find additional metadata (coordinates, crs, timestamp) in *metadata.parquet*. (For now, they are not automatically loaded)\n\n### Dataset opening\n\nUsing the dataset library, you download only the pre-patched raw version for simplicity.\n```python\nfrom dataset import load_dataset\n\n# There are two available configurations, \"post-fire\" and \"pre-post-fire.\"\ndataset = load_dataset(\"DarthReca/california_burned_areas\", name=\"post-fire\")\n```\n\nThe dataset was compressed using `h5py` and BZip2 from `hdf5plugin`. **WARNING: `hdf5plugin` is necessary to extract data**.\n\n### Data Instances\n\nEach matrix has a shape of 5490x5490xC, where C is 12 for pre-fire and post-fire images, while it is 0 for binary masks.\nPre-patched version is provided with matrices of size 512x512xC, too. In this case, only mask with at least one positive pixel is present.\n\nYou can find two versions of the dataset: _raw_ (without any transformation) and _normalized_ (with data normalized in the range 0-255). \nOur suggestion is to use the _raw_ version to have the possibility to apply any wanted pre-processing step.\n\n### Data Fields\n\nIn each standard HDF5 file, you can find post-fire, pre-fire images, and binary masks. The file is structured in this way:\n\n```bash\n├── foldn\n│   ├── uid0\n│   │   ├── pre_fire\n│   │   ├── post_fire\n│   │   ├── mask \n│   ├── uid1\n│       ├── post_fire\n│       ├── mask\n│  \n├── foldm\n    ├── uid2\n    │   ├── post_fire\n    │   ├── mask \n    ├── uid3\n        ├── pre_fire\n        ├── post_fire\n        ├── mask\n...\n```\n\nwhere `foldn` and `foldm` are fold names and `uidn` is a unique identifier for the wildfire.\n\nFor the pre-patched version, the structure is:\n```bash\nroot\n|\n|-- uid0_x: {post_fire, pre_fire, mask}\n|\n|-- uid0_y: {post_fire, pre_fire, mask}\n|\n|-- uid1_x: {post_fire, mask}\n|\n...\n```\nthe fold name is stored as an attribute.\n\n### Data Splits\n\nThere are 5 random splits whose names are: 0, 1, 2, 3, and 4.\n\n### Source Data\n\nData are collected directly from Copernicus Open Access Hub through the API. The band files are aggregated into one single matrix.\n\n## Additional Information\n\n### Licensing Information\n\nThis work is under the CDLA-permissive-2.0 license.\n\n### Citation Information\n\nIf you plan to use this dataset in your work please give the credit to Sentinel-2 mission and the California Department of Forestry and Fire Protection and cite using this BibTex:\n```\n@ARTICLE{cabuar,\n  author={Cambrin, Daniele Rege and Colomba, Luca and Garza, Paolo},\n  journal={IEEE Geoscience and Remote Sensing Magazine}, \n  title={CaBuAr: California burned areas dataset for delineation [Software and Data Sets]}, \n  year={2023},\n  volume={11},\n  number={3},\n  pages={106-113},\n  doi={10.1109/MGRS.2023.3292467}\n}\n```","encoding":"utf8","key":"README.md","name":"README.md","size":3846,"truncated":false}],"sourceObjectKey":"metadata.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"DarthReca/california_burned_areas\")\nprint(ds)","starterSamplePath":"metadata.parquet","totalObjects":45},"summary":"CaBuAr (California Burned Areas) is a burned-area segmentation dataset of Sentinel-2 scenes acquired before and after wildfires in California, with binary ground-truth masks…","title":"CaBuAr: California Burned Areas, Pre- and Post-Fire Sentinel-2 Scenes with Cal Fire Masks","updatedAt":"2026-11-17T20:20:43.847Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/DarthReca/california_burned_areas"}}