{"success":true,"data":{"author":"Orion Lab, National Observatory of Athens (SeasFire project)","bucketAccessType":"public","bucketName":"orion-ai-lab/seasfire_monthly","bucketPrefix":"","category":"Climate & Weather","cloudProvider":"huggingface","createdAt":"2026-11-16T14:11:25.930Z","description":"A machine-learning-ready global datacube for seasonal wildfire forecasting, derived from the SeasFire v0.4 cube (Zenodo 7108392) and covering 2001 to 2021. The source cube combines satellite, meteorological, climatological and human-influence variables relevant to fire, including burned area targets, on a 0.25 degree grid at 8-day steps (see the v0.4 record for the variable list). This release aggregates it to monthly steps, adds SPEI drought indices at several timescales, regional masks for Greece, California and New South Wales, and SatCLIP location embeddings. Files are ZIP-compressed Zarr stores: seasfire_orora_v0.1.zip (15.8 GB, 0.25 degree) and a 1 degree version (1.3 GB), SatCLIP embedding stores and their first five principal components, and under original_cube/ the unaggregated 8-day cube (43.9 GB at 0.25 degree, 2.7 GB at 1 degree). Open with xarray.open_zarr through a zip store. Released under CC BY 4.0 by the SeasFire team; cite the SeasFire datacube paper and the Orion Lab dataset.","displayTitle":"SeasFire Monthly: Global Seasonal Wildfire Forecasting Datacube, 2001-2021 (0.25 and 1 degree)","id":"ac184507-0ce4-5f06-9168-efeea0dc3b6e","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/orion-ai-lab/seasfire_monthly","objectCountEstimate":1082,"probeStatus":"ok","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"orion-ai-lab/seasfire_monthly\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"orion-ai-lab/seasfire_monthly\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=orion-ai-lab, license:cc-by-4.0, size_categories:10B<n<100B, downloads=1452"],"format":"mixed","generatedAt":"2026-11-16T14:11:33.543Z","layout":{"example":"original_cube/seasfire_1deg_v0.4.zip","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"original_cube/seasfire_1deg_v0.4.zip","truncatedAt":512}],"sampleObjects":["original_cube/seasfire_1deg_v0.4.zip","original_cube/seasfire_v0.4.zip","satclip-embedding.zip","satclip-embedding_1deg.zip","satclip_pcs_v0.1.zip"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\ntask_categories:\n- time-series-forecasting\ntags:\n- wildfire\n- earth-observation\n- climate\n- satellite\n- seasonal-forecasting\n- zarr\n- xarray\nsize_categories:\n- 10B<n<100B\nlanguage:\n- en\n---\n\n# SeasFire monthly: Seasonal Fire Forecasting Datacube (Monthly Resolution)\n\n## Dataset Description\n\nThe SeasFire monthly datacube is a comprehensive Earth observation dataset designed for **seasonal wildfire forecasting** using machine learning. It combines nearly 20 years (2001-2021) of multi-source satellite, meteorological, climatological, and human influence data into a unified, analysis-ready format.\n\nThis dataset is the result of extensive preprocessing and harmonization of the original [SeasFire v0.4 datacube](https://zenodo.org/records/7108392), with additional features including:\n- Monthly temporal aggregation for seasonal forecasting tasks\n- Integrated drought indices (SPEI at multiple timescales)\n- Regional masks for targeted analysis (Greece, California, New South Wales)\n- SATCLIP satellite image embeddings for enhanced feature representation\n\n### Dataset Summary\n\n- **Temporal Coverage**: 2001-2021 (20 years)\n- **Temporal Resolution**: Monthly\n- **Spatial Resolutions**: 0.25° (~25km)\n- **Format**: Zarr (ZIP-compressed for portability)\n- **Variables**: Refer to [SeasFire v0.4 datacube](https://zenodo.org/records/7108392) for comprehensive variable analysis.\n- **Use Cases**: Wildfire prediction, seasonal forecasting, climate analysis, Earth system modeling\n\n## Available Files\n\n### Preprocessed Datacubes (Ready to Use)\n\n| File | Size | Resolution | Description |\n|------|------|------------|-------------|\n| `seasfire_orora_v0.1.zip` | 15.8 GB | 0.25° spatial, monthly temporal | Main datacube with all features |\n| `seasfire_orora_1deg_v0.1.zip` | 1.26 GB | 1° spatial, monthly temporal | Coarsened version for faster processing |\n\n**These are the recommended files for most users.** They include monthly aggregation, drought indices, regional masks, and metadata.\n\n### SATCLIP Embeddings (Optional Enhancement)\n\n| File | Size | Resolution | Description |\n|------|------|------------|-------------|\n| `satclip-embedding.zip` | 3.78 GB | 0.25° | Full SATCLIP embeddings from satellite imagery |\n| `satclip-embedding_1deg.zip` | 243 MB | 1° | Coarsened SATCLIP embeddings |\n| `satclip_pcs_v0.1.zip` | 163 MB | 0.25° | Principal components (first 5) of SATCLIP embeddings |\n\n**SATCLIP embeddings** are learned representations from satellite images using contrastive learning. They can enhance model performance but are optional.\n\n### Original SeasFire Cube\n\n| File | Size | Resolution | Description |\n|------|------|------------|-------------|\n| `original_cube/seasfire_v0.4.zip` | 43.9 GB | 0.25° spatial, 8-daily temporal | Original SeasFire datacube before preprocessing |\n| `original_cube/seasfire_1deg_v0.4.zip` | 2.69 GB | 1° spatial, 8-daily temporal | Coarsened original datacube |\n\n**Use these only if** you need the original 8-daily temporal resolution or want to customize the preprocessing pipeline.\n\n## Additional Data Variables\n\n### Drought Indices (SPEI)\n- **spei_1, spei_3, spei_6, spei_12, spei_24, spei_36, spei_48**: Standardized Precipitation-Evapotranspiration Index at multiple timescales (1-48 months)\n\n### SATCLIP Features (in separate files)\n- **satclip_embeddings**: Learned representations from satellite imagery\n- **satclip_embeddings_pc1 to pc5**: Principal components of embeddings\n\n## Loading the Data\n\n### Python with xarray\n\n```python\nimport xarray as xr\n\n# Load the main datacube\nds = xr.open_zarr(\"seasfire_orora_v0.1.zip\", consolidated=True)\n\n# Or load the 1-degree version for faster processing\nds = xr.open_zarr(\"seasfire_orora_1deg_v0.1.zip\", consolidated=True)\n\n# Explore the dataset\nprint(ds)\nprint(ds.data_vars)\n\n# Select a specific time period\nds_subset = ds.sel(time=slice(\"2020-01-01\", \"2021-12-31\"))\n\n# Select a specific region (e.g., California)\ncalifornia_data = ds.where(ds.regions_of_interest == 2, drop=False)\n```\n\n### Downloading with Hugging Face Hub\n\n```python\nfrom huggingface_hub import h","encoding":"utf8","key":"README.md","name":"README.md","size":5446,"truncated":true}],"sourceObjectKey":"original_cube/seasfire_1deg_v0.4.zip","starterCell":"from huggingface_hub import hf_hub_download\npath = hf_hub_download(repo_id=\"orion-ai-lab/seasfire_monthly\", filename=\"original_cube/seasfire_1deg_v0.4.zip\", repo_type=\"dataset\")\nprint(path)","starterSamplePath":"original_cube/seasfire_1deg_v0.4.zip","totalObjects":1082},"summary":"A machine-learning-ready global datacube for seasonal wildfire forecasting, derived from the SeasFire v0.4 cube (Zenodo 7108392) and covering 2001 to 2021.","title":"SeasFire Monthly: Global Seasonal Wildfire Forecasting Datacube, 2001-2021 (0.25 and 1 degree)","updatedAt":"2026-11-16T16:15:44.578Z","visibility":"public","websiteUrl":"https://seasfire.hua.gr/"}}