{"success":true,"data":{"author":"EPFL Environmental Computational Science and Earth Observation Laboratory (ECEO)","bucketAccessType":"public","bucketName":"EPFL-ECEO/CanadaFireSat","bucketPrefix":"","category":"Environmental Science","cloudProvider":"huggingface","createdAt":"2026-11-14T14:15:15.865Z","description":"Benchmark for high-resolution (100 m) wildfire occurrence forecasting across Canada with multimodal deep learning. Fire targets come from the National Burned Area Composite (NBAC) polygons, mapped to a 2.8 km grid in 8-day windows. Inputs for each sample are a Sentinel-2 Level-1C satellite image time series over the 64 days before prediction (cloud cover under 40 percent) plus environmental predictors: ERA5-Land hydrometeorology, MODIS land surface temperature, MODIS NDVI, EVI, LAI and FPAR, and Canadian Fire Weather Index and Drought Code from Copernicus EMS. The card reports 177,801 samples from 2016 to 2023, with fire in 39 percent of samples, split by year (78,030 train 2016 to 2021, 14,329 validation 2022, 85,442 test 2023), and NALCMS land cover for analysis. The listing (measured 2026-11-14) shows one folder per province or territory (Alberta alone holds 502 Parquet shards, about 182 GB) plus images/ and metadata/; the repository uses about 1.16 TB. Load with Hugging Face datasets or pyarrow. CNN and ViT baselines are on GitHub; an NPY version is EPFL-ECEO/CanadaFireSat-Raw. Released under the MIT licence; cite the ISPRS paper (arXiv:2506.08690).","displayTitle":"CanadaFireSat: High-Resolution Multimodal Wildfire Forecasting Benchmark for Canada","id":"54f83cdd-9796-55a4-b7f0-d84e2f8219b8","license":"MIT","licenseEvidenceUrl":"https://huggingface.co/datasets/EPFL-ECEO/CanadaFireSat","objectCountEstimate":3003,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2506.08690","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"EPFL-ECEO/CanadaFireSat\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"EPFL-ECEO/CanadaFireSat\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=EPFL-ECEO, license:mit, size_categories:100K<n<1M, format:parquet, downloads=4805"],"format":"mixed","generatedAt":"2026-11-14T14:15:22.584Z","layout":{"example":"Alberta/test-00000-of-00208.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"Alberta/test-00000-of-00208.parquet","truncatedAt":512}],"sampleObjects":["Alberta/test-00000-of-00208.parquet","Alberta/test-00001-of-00208.parquet","Alberta/test-00002-of-00208.parquet","Alberta/test-00003-of-00208.parquet","Alberta/test-00004-of-00208.parquet"],"sidecars":[{"bytes":"---\ndataset_info:\n- config_name: Alberta\n  features:\n  - name: date\n    dtype: timestamp[s]\n  - name: doy\n    sequence: int64\n  - name: 10x\n    sequence:\n      array3_d:\n        shape:\n        - 4\n        - 264\n        - 264\n        dtype: uint8\n  - name: 20x\n    sequence:\n      array3_d:\n        shape:\n        - 6\n        - 132\n        - 132\n        dtype: uint8\n  - name: 60x\n    sequence:\n      array3_d:\n        shape:\n        - 3\n        - 44\n        - 44\n        dtype: uint8\n  - name: loc\n    dtype:\n      array3_d:\n        shape:\n        - 2\n        - 264\n        - 264\n        dtype: float32\n  - name: labels\n    dtype:\n      array2_d:\n        shape:\n        - 264\n        - 264\n        dtype: uint8\n  - name: tab_cds\n    dtype:\n      array2_d:\n        shape:\n        - 8\n        - 6\n        dtype: float32\n  - name: tab_era5\n    dtype:\n      array2_d:\n        shape:\n        - 8\n        - 45\n        dtype: float32\n  - name: tab_modis\n    dtype:\n      array2_d:\n        shape:\n        - 8\n        - 7\n        dtype: float32\n  - name: env_cds\n    dtype:\n      array4_d:\n        shape:\n        - 8\n        - 6\n        - 13\n        - 13\n        dtype: float32\n  - name: env_cds_loc\n    dtype:\n      array3_d:\n        shape:\n        - 13\n        - 13\n        - 2\n        dtype: float32\n  - name: env_era5\n    dtype:\n      array4_d:\n        shape:\n        - 8\n        - 45\n        - 32\n        - 32\n        dtype: float32\n  - name: env_era5_loc\n    dtype:\n      array3_d:\n        shape:\n        - 32\n        - 32\n        - 2\n        dtype: float32\n  - name: env_modis11\n    dtype:\n      array4_d:\n        shape:\n        - 8\n        - 3\n        - 16\n        - 16\n        dtype: float32\n  - name: env_modis11_loc\n    dtype:\n      array3_d:\n        shape:\n        - 16\n        - 16\n        - 2\n        dtype: float32\n  - name: env_modis13_15\n    dtype:\n      array4_d:\n        shape:\n        - 8\n        - 4\n        - 32\n        - 32\n        dtype: float32\n  - name: env_modis13_15_loc\n    dtype:\n      array3_d:\n        shape:\n        - 32\n        - 32\n        - 2\n        dtype: float32\n  - name: env_doy\n    sequence: int64\n  - name: region\n    dtype: string\n  - name: tile_id\n    dtype: int32\n  - name: file_id\n    dtype: string\n  - name: fwi\n    dtype: float32\n  splits:\n  - name: train\n    num_bytes: 55480836258\n    num_examples: 9345\n  - name: validation\n    num_bytes: 10599466718\n    num_examples: 1727\n  - name: test\n    num_bytes: 103887942242\n    num_examples: 16937\n  - name: test_hard\n    num_bytes: 80360022890\n    num_examples: 12145\n  download_size: 182146467041\n  dataset_size: 250328268108\n- config_name: British Columbia\n  features:\n  - name: date\n    dtype: timestamp[s]\n  - name: doy\n    sequence: int64\n  - name: 10x\n    sequence:\n      array3_d:\n        shape:\n        - 4\n        - 264\n        - 264\n        dtype: uint8\n  - name: 20x\n    sequence:\n      array3_d:\n        shape:\n        - 6\n        - 132\n        - 132\n        dtype: uint8\n  - name: 60x\n    sequence:\n      array3_d:\n        shape:\n        - 3\n        - 44\n        - 44\n        dtype: uint8\n  - name: loc\n    dtype:\n      array3_d:\n        shape:\n        - 2\n        - 264\n        - 264\n        dtype: float32\n  - name: labels\n    dtype:\n      array2_d:\n        shape:\n        - 264\n        - 264\n        dtype: uint8\n  - name: tab_cds\n    dtype:\n      array2_d:\n        shape:\n        - 8\n        - 6\n        dtype: float32\n  - name: tab_era5\n    dtype:\n      array2_d:\n        shape:\n        - 8\n        - 45\n        dtype: float32\n  - name: tab_modis\n    dtype:\n      array2_d:\n        shape:\n        - 8\n        - 7\n        dtype: float32\n  - name: env_cds\n    dtype:\n      array4_d:\n        shape:\n        - 8\n        - 6\n        - 13\n        - 13\n        dtype: float32\n  - name: env_cds_loc\n    dtype:\n      array3_d:\n        shape:\n        - 13\n        - 13\n        - 2\n        dtype: float32\n  - name: env_era5\n    dtype:\n      array4_d:\n        shape:\n        - 8\n        - 45\n        - 32\n        - 32\n        dtype: float32\n  - name: env_era5_loc\n    dtype:\n      array3_d:\n        shap","encoding":"utf8","key":"README.md","name":"README.md","size":45941,"truncated":true}],"sourceObjectKey":"Alberta/test-00000-of-00208.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"EPFL-ECEO/CanadaFireSat\")\nprint(ds)","starterSamplePath":"Alberta/test-00000-of-00208.parquet","totalObjects":3003},"summary":"Benchmark for high-resolution (100 m) wildfire occurrence forecasting across Canada with multimodal deep learning.","title":"CanadaFireSat: High-Resolution Multimodal Wildfire Forecasting Benchmark for Canada","updatedAt":"2026-11-14T17:43:45.147Z","visibility":"public","websiteUrl":"https://github.com/eceo-epfl/CanadaFireSat-Data"}}