{"success":true,"data":{"author":"Liangyun Liu, Xiao Zhang and Zhehua Li (Aerospace Information Research Institute, Chinese Academy of Sciences)","bucketAccessType":"public","bucketListable":true,"bucketName":"us-west-2.opendata.source.coop","bucketPrefix":"portolan-mirrors/glc-fcs30d","category":"Earth Observation & Geospatial","cloudProvider":"aws","createdAt":"2026-11-21T20:46:03.409Z","description":"GLC_FCS30D version 2 is a global land-cover map at 30 m resolution with a fine classification system of 35 land-cover categories, for 1985, 1990, 1995 and every year from 2000 to 2022, made by Liangyun Liu, Xiao Zhang and Zhehua Li at AIRCAS (Chinese Academy of Sciences) from Landsat time series with continuous change detection. The producer reports overall accuracy of 80.88% for the basic 10-class system and 73.24% for the 17 LCCS level-1 classes. Use it to measure forest loss, cropland expansion, urban growth (class 190, impervious surface) or any land-cover transition over four decades. The mirror holds 24,986 single-band uint8 Cloud Optimized GeoTIFFs, one per 5 x 5 degree tile (961 tiles) per year, in EPSG:4326 with embedded color tables and nodata 0, grouped by year and 10-degree longitude band under glc_fcs30d/, plus one STAC item per COG and a stac-geoparquet items.parquet for finding tiles by point and year with DuckDB. Read with GDAL, rasterio or rioxarray. Pixel values are unchanged from the Zenodo release. Licensed CC BY 4.0 on the Zenodo record (DOI 10.5281/zenodo.15063683); cite Zhang et al. 2024, Earth System Science Data 16, 1353, doi:10.5194/essd-16-1353-2024.","displayTitle":"GLC_FCS30D v2: Global 30 m Land Cover Dynamics with 35 classes, 1985-2022 (AIRCAS, Source Cooperative)","doi":"10.5281/zenodo.15063683","fuseMountBrowsable":true,"id":"3c9e8893-a3d8-5737-8c1d-dbdbb2757e43","license":"CC-BY-4.0","licenseEvidenceUrl":"https://zenodo.org/records/15063683","objectCountEstimate":1,"probeStatus":"ok","referenceUrl":"https://doi.org/10.5194/essd-16-1353-2024","region":"us-west-2","schemaCard":{"accessHints":[{"language":"python-anonymous","snippet":"import boto3\nfrom botocore import UNSIGNED\nfrom botocore.client import Config\ns3 = boto3.client('s3', region_name='us-west-2', config=Config(signature_version=UNSIGNED))\ns3.list_objects_v2(Bucket='us-west-2.opendata.source.coop', Prefix='portolan-mirrors/glc-fcs30d/', Delimiter='/')"},{"language":"python-local","snippet":"open('/mnt/archive/glc_fcs30d/1985/AGENTS.md', 'rb')"}],"caveats":["Files are binary — use a domain-specific parser (e.g. metpy / pynexrad / pydicom) rather than text I/O."],"format":"binary","generatedAt":"2026-11-21T20:48:56.941Z","layout":{"example":"portolan-mirrors/glc-fcs30d/glc_fcs30d/1985/AGENTS.md","pattern":"{prefix}/{seg1}/{seg2}/{YYYY}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"# Agent guide: GLC_FCS30D 1985\n\n- This catalog holds 36 band catalogs and 961 items, one for each 5° tile. Each item has one single-band uint8 COG of class values in EPSG:4326, nodata 0.\n- To find the tiles for a place, query `https://data.source.coop/portolan-mirrors/glc-fcs30d/glc_fcs30d/items.parquet` with DuckDB and filter on `bbox` and `start_datetime`, instead of walking these catalogs.\n- Read the [collection agent guide](../AGENTS.md) before you compute areas or compare years.\n","encoding":"utf8","key":"portolan-mirrors/glc-fcs30d/glc_fcs30d/1985/AGENTS.md","truncatedAt":512},{"bytes":"{\n  \"type\": \"Feature\",\n  \"stac_version\": \"1.1.0\",\n  \"stac_extensions\": [\n    \"https://stac-extensions.github.io/file/v2.1.0/schema.json\",\n    \"https://stac-extensions.github.io/projection/v2.0.0/schema.json\",\n    \"https://stac-extensions.github.io/classification/v2.0.0/schema.json\"\n  ],\n  \"id\": \"1985-E0N10\",\n  \"collection\": \"glc_fcs30d\",\n  \"geometry\": {\n    \"type\": \"Polygon\",\n    \"coordinates\": [\n      [\n        [\n          0.0,\n          4.999797466\n        ],\n        [\n          5.000202534,\n          4.9","encoding":"utf8","key":"portolan-mirrors/glc-fcs30d/glc_fcs30d/1985/E0-E5/1985-E0N10/1985-E0N10.json","truncatedAt":512}],"sampleObjects":["portolan-mirrors/glc-fcs30d/glc_fcs30d/1985/AGENTS.md","portolan-mirrors/glc-fcs30d/glc_fcs30d/1985/E0-E5/1985-E0N10/1985-E0N10.json","portolan-mirrors/glc-fcs30d/glc_fcs30d/1985/E0-E5/1985-E0N10/1985-E0N10.tif","portolan-mirrors/glc-fcs30d/glc_fcs30d/1985/E0-E5/1985-E0N15/1985-E0N15.json","portolan-mirrors/glc-fcs30d/glc_fcs30d/1985/E0-E5/1985-E0N15/1985-E0N15.tif"],"sidecars":[{"bytes":"# GLC_FCS30D: Global 30 m Land Cover Dynamics, 1985–2022\n\nA cloud-native mirror of [GLC_FCS30D v2](https://doi.org/10.5281/zenodo.15063683): land cover of the whole world at 30 m, in 35 classes. There are maps for 1985, 1990, 1995 and every year from 2000 to 2022.\n\n## What is here\n\nOne collection, [`glc_fcs30d`](./glc_fcs30d/README.md):\n- **24,986 maps.** Each is a Cloud Optimized GeoTIFF: one for each of 961 tiles of 5° by 5°, for each of 26 years. Each COG has the class colors in its color table.\n- **One STAC item per COG.** The items are grouped by year, then by 10° longitude band.\n- **[`items.parquet`](./glc_fcs30d/items.parquet).** Every item in one stac-geoparquet file, for search.\n\n## License\n\n[CC BY 4.0](https://creativecommons.org/licenses/by/4.0/), as given on the Zenodo record. The rights holders are the producers: Liangyun Liu, Xiao Zhang and Zhehua Li, Aerospace Information Research Institute, Chinese Academy of Sciences. Cite [doi:10.5194/essd-16-1353-2024](https://doi.org/10.5194/essd-16-1353-2024).\n\n## Provenance\n\nThis is a mirror, not the official publication. The upstream is the Zenodo record [doi:10.5281/zenodo.15063683](https://doi.org/10.5281/zenodo.15063683), version v2 of 2025-03-21. The mirror was synced once. The `updated` field on this catalog records the time of that sync. The pixel values are unchanged. Only the file layout changed: one COG per tile and year, instead of one multi-year GeoTIFF per tile inside 36 ZIP files. The tools and the metadata are in <https://github.com/portolan-mirrors/glc-fcs30d>.\n\n## Access\n\nFind the 2022 tile that holds a point, with no download:\n\n```sql\nSELECT id, assets.data.href\nFROM 'https://data.source.coop/portolan-mirrors/glc-fcs30d/glc_fcs30d/items.parquet'\nWHERE start_datetime >= TIMESTAMP '2022-01-01' AND start_datetime < TIMESTAMP '2023-01-01'\n  AND bbox.xmin <= 2.35 AND bbox.xmax >= 2.35 AND bbox.ymin <= 48.86 AND bbox.ymax >= 48.86;\n```\n\nThen open the `href` with GDAL as `/vsicurl/<href>`. See the [collection README](./glc_fcs30d/README.md) for the classes and the known issues.\n","encoding":"utf8","key":"portolan-mirrors/glc-fcs30d/README.md","name":"README.md","size":2085,"truncated":false},{"bytes":"{\n  \"type\": \"Catalog\",\n  \"stac_version\": \"1.1.0\",\n  \"stac_extensions\": [\n    \"https://schemas.portolan-sdi.org/portolan/v0.3.0/schema.json\"\n  ],\n  \"id\": \"glc-fcs30d\",\n  \"title\": \"GLC_FCS30D: Global 30 m Land Cover Dynamics, 1985–2022\",\n  \"description\": \"A cloud-native mirror of [GLC_FCS30D v2](https://doi.org/10.5281/zenodo.15063683), a global 30 m land cover time series with 35 classes, produced by Liangyun Liu, Xiao Zhang and Zhehua Li at the Aerospace Information Research Institute, Chinese Academy of Sciences. It has one map every 5 years from 1985 to 1995 and one map each year from 2000 to 2022, as Cloud Optimized GeoTIFFs with embedded class colors, one for each 5° tile and year. Agents start at [AGENTS.md](AGENTS.md).\",\n  \"links\": [\n    {\n      \"rel\": \"self\",\n      \"href\": \"https://data.source.coop/portolan-mirrors/glc-fcs30d/catalog.json\",\n      \"type\": \"application/json\"\n    },\n    {\n      \"rel\": \"root\",\n      \"href\": \"./catalog.json\",\n      \"type\": \"application/json\"\n    },\n    {\n      \"rel\": \"describedby\",\n      \"href\": \"./README.md\",\n      \"type\": \"text/markdown\",\n      \"title\": \"Catalog README\"\n    },\n    {\n      \"rel\": \"agents\",\n      \"href\": \"./AGENTS.md\",\n      \"type\": \"text/markdown\",\n      \"title\": \"Catalog agent guide\"\n    },\n    {\n      \"rel\": \"alternate\",\n      \"href\": \"https://browser.portolan-sdi.org/#/external/data.source.coop/portolan-mirrors/glc-fcs30d/catalog.json\",\n      \"type\": \"text/html\",\n      \"title\": \"Open in the data browser\"\n    },\n    {\n      \"rel\": \"vcs\",\n      \"href\": \"https://github.com/portolan-mirrors/glc-fcs30d\",\n      \"type\": \"text/html\",\n      \"title\": \"Source repository\"\n    },\n    {\n      \"rel\": \"issues\",\n      \"href\": \"https://github.com/portolan-mirrors/glc-fcs30d/issues\",\n      \"type\": \"text/html\",\n      \"title\": \"Report a problem\"\n    },\n    {\n      \"rel\": \"about\",\n      \"href\": \"https://source.coop/portolan-mirrors/glc-fcs30d\",\n      \"type\": \"text/html\",\n      \"title\": \"Hosting page on Source Cooperative\"\n    },\n    {\n      \"rel\": \"child\",\n      \"href\": \"./glc_fcs30d/collection.json\",\n      \"type\": \"application/json\",\n      \"title\": \"GLC_FCS30D global 30 m land cover, 1985–2022\"\n    }\n  ],\n  \"updated\": \"2026-11-20T17:05:25Z\"\n}\n","encoding":"utf8","key":"portolan-mirrors/glc-fcs30d/catalog.json","name":"catalog.json","size":2222,"truncated":false}],"sourceObjectKey":"portolan-mirrors/glc-fcs30d/glc_fcs30d/1985/AGENTS.md","starterCell":"with open('/mnt/archive/glc_fcs30d/1985/AGENTS.md', 'rb') as fh:\n    head = fh.read(1024)\nprint(len(head), head[:64])","starterSamplePath":"glc_fcs30d/1985/E0-E5/1985-E0N10/1985-E0N10.tif","totalObjects":"1 top-level prefixes"},"summary":"GLC_FCS30D version 2 is a global land-cover map at 30 m resolution with a fine classification system of 35 land-cover categories, for 1985, 1990, 1995 and every year from 2000 to…","title":"GLC_FCS30D v2: Global 30 m Land Cover Dynamics with 35 classes, 1985-2022 (AIRCAS, Source Cooperative)","updatedAt":"2026-11-21T22:20:49.223Z","visibility":"public","websiteUrl":"https://source.coop/portolan-mirrors/glc-fcs30d"}}