Modeled 1 km allocation, not observed livestock locations: The pixels represent a statistical redistribution of coarse subnational census counts across annual layer of potential land for livestock production; they do not mark observed herds or farms. Model accuracy was evaluated on held-out census polygons using polygon-mean covariates. As the mean census unit is ~2,900 km\u00b2 (SD ~21,500), these metrics describe agreement at the administrative-unit level and do not characterize the accuracy of the 1 km detail. Treat the fine-scale pattern as a plausible disaggregation, not an independently validated 1 km estimate.
FAOSTAT consistency is not local accuracy: The country-by-country rescaling guarantees that national totals match FAOSTAT by construction; it does not constrain or validate where within a country animals are placed. National totals can therefore be correct while subnational allocation remains uncertain.
Discontinuities at national borders: Because headcounts are rescaled to FAOSTAT national totals separately for each country, abrupt changes can appear across international boundaries \u2014 for example between Mongolia and its neighbours, where reported national totals differ strongly. These are artifacts of the adjustment, not real density gradients.
Irregular and incomplete census data: Livestock census data are often irregular, incomplete, and can be overestimated due to double counting caused by animal mobility across census boundaries or changes in farm ownership. Furthermore, spatial availability varies widely; many developing countries only provide data at coarse administrative scales (state/county level), which may lead to systematic underestimation of undocumented livestock.
Mismatched input scales and unrealistic densities: Livestock densities are derived by combining administrative headcount estimates with the annual layer of potential land for livestock production. In regions dominated by landless livestock systems or feedlots, the high number of livestock may not match the mapped potential land, resulting in unrealistically high density values (locally exceeding 2,500 cattle heads km\u207b\u00b2).
Uniform allocation across livestock species: The intermediate layer representing potential land for livestock production is applied uniformly as an input for all modeled livestock species. The current model does not account for the complexities of different livestock management systems (dairy vs. cow-calf vs. finishing systems) or distinct species-specific behaviors, such as goats acting primarily as browsers.
Underestimation of high-density areas and unreliable prediction intervals: The selected machine learning models struggle to predict areas with very high livestock density, likely underestimating intense livestock hotspots. The prediction intervals are much narrower than their nominal 95% level implies and understate the true uncertainty. Although labeled as 95% intervals, Prediction Interval Coverage Probability (PICP) was found to be 30-45%, meaning that held-out census-polygon observations fell inside the prediction intervals only about 30-45% of the time. These intervals should not be used as probabilistic bounds for risk thresholds, exceedance probabilities, or downstream error propagation without independent recalibration.
Temporal coverage vs. validated dynamics: The layers span 2000\u20132022 annually, but there is no temporal hold-out validation. Much of the inter-annual signal derives from FAOSTAT national totals and the annually varying land-cover inputs, while the per-pixel density model relies substantially on static and long-term covariates; some socioeconomic inputs are held constant and carried forward (e.g., HDI to 2015, nighttime lights to ~2020). Interpret short-term (year-to-year) change cautiously.
Interrupted Goode Homolosine projection caveats: To ensure precise spatial alignment and minimize distortions, all modeling was conducted in an equal-area coordinate system, specifically the Interrupted Goode Homolosine projection (used for all layers in Zenodo). Prior to GEE ingestion, the dataset was converted to the EPSG:4326 coordinate system. Because the original headcount data represents absolute values per 1 km\u00b2, a spatial redistribution approach was used to correct for projection-induced distortions (see Github). This adjustment preserves total national values for consistency with FAOSTAT statistics, though it may slightly shift the local allocation of headcounts within country borders.
", "citation": "", "creator_user_id": "b1d09433-9a11-4bb5-b8ba-918ddd9814fa", "draft": false, "featured_dataset": false, "has_chart_views": false, "id": "985d8f45-6e25-480b-9d04-af77b55ba66e", "is_approved": true, "isopen": true, "language": "en", "learn_more": "https://landcarbonlab.org/insights/global-livestock-dynamics", "license_id": "cc-by", "license_title": "Creative Commons Attribution", "license_url": "http://www.opendefinition.org/licenses/cc-by", "maintainer": null, "maintainer_email": null, "maintainers": [{"name": "Radost Stanimirova", "email": "radost.stanimirova@wri.org"}], "metadata_created": "2026-03-02T15:59:40.902809", "metadata_modified": "2026-07-17T14:52:03.271172", "methodology": "We used an areal regression, spatiotemporal machine-learning framework to produce annual 1 km maps of livestock densities and headcounts for cattle, horses, sheep, goats, and buffaloes from 2000\u20132022 . Subnational census data from 55,336 administrative units across 147 countries were harmonized and linked to 128 environmental and socio-economic raster predictors, with livestock densities calculated using a dynamic suitability mask derived from annual grassland and cropland extent and constrained by temperature thresholds. Species-specific Random Forest models were optimized using recursive feature elimination and hyperparameter tuning, and evaluated with spatially blocked training, calibration, and testing splits (200 \u00d7 200 km blocks) to reduce spatial autocorrelation bias . Performance was assessed using metrics appropriate for skewed, Poisson-like data, including the Poisson deviance\u2013based D\u00b2 score and Concordance Correlation Coefficient . Final 1 km predictions include 95% prediction intervals derived from ensemble quantiles and are scaled using country-level adjustment factors so that aggregated grid totals match FAOSTAT national statistics, ensuring consistency with official reporting systems .
Link to paper: https://peerj.com/articles/21494/
Link to Zenodo: https://zenodo.org/records/20396303
Link to GitHub: global-pasture-watch/gld-1km at main \u00b7 wri/global-pasture-watch
Link to STAC: https://browser.stac.opengeohub.org/cat/landmetric?.language=en
", "name": "annual-livestock-headcount", "notes": "These products provide the first globally consistent, annually updated (2000\u20132022) 1-km resolution maps of cattle, sheep, goats, horses, and buffalo densities and headcounts, addressing a critical gap in high-resolution livestock data needed for climate mitigation, land-use planning, biodiversity conservation, and food systems analysis. Livestock systems are central to global food security and rural livelihoods, yet they are also major drivers of greenhouse gas emissions, deforestation, land degradation, and biodiversity loss; effective policy and modeling require spatially explicit, temporally dynamic data that go beyond coarse national statistics or 5\u201310 km gridded products . Using the largest harmonized compilation of subnational census data to date (55,336 administrative units across 147 countries) combined with 128 environmental and socio-economic raster layers, we applied a spatiotemporal machine learning framework (Random Forest with spatial blocking and feature selection) to model annual livestock densities, which were subsequently adjusted to align with FAOSTAT national totals . The resulting outputs include annual 1-km maps of livestock densities with 95% prediction intervals and FAOSTAT-consistent headcounts, enabling both uncertainty-aware analysis and compatibility with official statistics . These data can and should be used to improve greenhouse gas inventories, land-use change assessments, grazing system analyses, and national policy planning, while users conducting subnational applications are encouraged to recalibrate headcounts using local census data where available and interpret prediction intervals to account for uncertainty.
As part of this work, we have generated four products:
Land suitable for livestock production - The annual 1 km modeling spatial domain representing the fraction of each grid cell derived from integrated grassland and cropland extent and constrained by temperature thresholds, used to allocate livestock densities only to areas biophysically capable of supporting grazing or forage systems. Land suitable for livestock production can be found here [Zenodo].
Livestock density \u2013 Modeled livestock density expressed as heads per km\u00b2 within the estimated fraction of land suitable for livestock production, derived through spatiotemporal machine learning. Livestock density per species can be found here [Zenodo].
Headcount (raw) \u2013 Grid-level livestock headcounts calculated by multiplying predicted densities by the fraction of potential livestock land in each 1 km cell prior to national calibration. Headcount (raw) per livestock species can be found here.
Headcount (FAOSTAT-adjusted) \u2013 Grid-level livestock headcounts scaled using country-specific adjustment factors so that the sum of all 1 km cells matches official FAOSTAT national totals for each species and year. FAOStat adjusted headcount per livestock species can be found here [Zenodo].
Here we show the FAOStat adjusted headcount data for cattle, but these data are available for four other livestock species: sheep, goats, horses, and buffaloes. You can find the headcount data for other livestock species on Zenodo using the following links:
The visualization shown here is for the year 2022, but the data is available from 2000-2022 in Zenodo.
", "num_resources": 6, "num_tags": 5, "open_in": "[{\"title\":\"Google Earth Engine\",\"url\":\"https://developers.google.com/earth-engine/datasets/catalog/projects_global-pasture-watch_assets_gld-1km_v1_livestock-headcount-faostat_m\"}]", "organization": {"id": "ad461dbd-f646-4080-94a9-c10005c42a43", "name": "land-carbon-lab", "title": "Land & Carbon Lab", "type": "organization", "description": "WRI's Land & Carbon Lab creates datasets to support critical monitoring and decision making needs surrounding land use planning, carbon accounting, and other technical facets of human-land interactions.", "image_url": "1724353785-cover-ethiopia-mosaic-oqiid6.avif", "created": "2024-09-25T11:07:47.708772", "is_organization": true, "approval_status": "approved", "state": "active"}, "owner_org": "ad461dbd-f646-4080-94a9-c10005c42a43", "private": false, "project": "", "rw_dataset": false, "rw_id": "", "short_description": "Subnational livestock census data and satellite based estimates of the distribution of cattle, horses, goats, sheep and buffaloes are well suited for understanding food systems, land use, and environmental impacts.", "spatial": "null", "spatial_address": "Global", "spatial_type": "global", "state": "active", "technical_notes": "https://peerj.com/articles/21494/", "temporal_coverage_end": "2022", "temporal_coverage_start": "2000", "title": "Global distribution of cattle, horses, goats, sheep and buffaloes at 1 km resolution for 2000-2022", "type": "dataset", "update_frequency": "annually", "version": null, "visibility_type": "public", "wri_data": true, "groups": [{"description": "Data concerning land use, land cover, and terrestrial ecosystem dynamics along with the human and environmental drivers of our food, forests, and water systems.", "display_name": "Land", "id": "c1427a08-9de5-4182-9c05-b6a61fda3127", "image_display_url": "https://datasets.wri.org/private-admin/uploads/group/land-peru-amazon-river-jrjo7f.jpg", "name": "land", "title": "Land", "type": "group"}], "resources": [{"cache_last_updated": null, "cache_url": null, "created": "2026-07-13T13:28:57.219000", "datastore_active": false, "description": "This data product has been archived with Zenodo and you are able to access the GeoTIFFs using this link.", "format": "COG", "hash": "", "id": "61f5f533-1b71-4fc3-ab0d-198c892ae612", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-07-17T14:52:03.281661", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": true, "package_id": "985d8f45-6e25-480b-9d04-af77b55ba66e", "position": 0, "resourceId": "61f5f533-1b71-4fc3-ab0d-198c892ae612", "resource_type": null, "schema": {"value": []}, "size": null, "state": "active", "title": "Cloud Optimized GeoTIFFs (COGs)", "type": "link", "url": "https://zenodo.org/records/17491242", "url_type": "link"}, {"cache_last_updated": null, "cache_url": null, "created": "2026-07-13T13:28:57.219000", "datastore_active": false, "description": "Cattle headcount for year 2022", "format": "", "hash": "", "id": "8ac0c160-d851-4e5d-bbf0-6475a005f230", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-07-17T14:52:03.281783", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "985d8f45-6e25-480b-9d04-af77b55ba66e", "position": 1, "resourceId": "8ac0c160-d851-4e5d-bbf0-6475a005f230", "resource_type": null, "rw_id": "b8c846ea-cd5a-43b2-abba-8e94ebe9507a", "schema": {"value": []}, "size": null, "state": "active", "title": "Cattle headcount in 2022", "type": "reference-layer", "url": "", "url_type": "reference-layer"}, {"cache_last_updated": null, "cache_url": null, "created": "2026-07-13T13:28:57.219000", "datastore_active": false, "description": "Sheep headcount for year 2022", "format": "", "hash": "", "id": "b0508a84-23d9-4f68-abd2-a1769f9fe031", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-07-17T14:52:03.281860", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "985d8f45-6e25-480b-9d04-af77b55ba66e", "position": 2, "resourceId": "b0508a84-23d9-4f68-abd2-a1769f9fe031", "resource_type": null, "rw_id": "7a00bbd8-9877-4ecc-97c4-d75316306375", "schema": {"value": []}, "size": null, "state": "active", "title": "Sheep headcount in 2022", "type": "reference-layer", "url": "", "url_type": "reference-layer"}, {"cache_last_updated": null, "cache_url": null, "created": "2026-07-13T13:28:57.219000", "datastore_active": false, "description": "Goat headcount for year 2022", "format": "", "hash": "", "id": "07463d26-daa8-431f-9262-dc1f308b24d3", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-07-17T14:52:03.281932", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "985d8f45-6e25-480b-9d04-af77b55ba66e", "position": 3, "resourceId": "07463d26-daa8-431f-9262-dc1f308b24d3", "resource_type": null, "rw_id": "078e7bd8-735d-4226-9093-4c15fdc025f3", "schema": {"value": []}, "size": null, "state": "active", "title": "Goat headcount in 2022", "type": "reference-layer", "url": "", "url_type": "reference-layer"}, {"cache_last_updated": null, "cache_url": null, "created": "2026-07-13T13:28:57.219000", "datastore_active": false, "description": "Buffalo headcount for year 2022", "format": "", "hash": "", "id": "fab57d92-2a95-4220-ba02-5082a8b6c028", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-07-17T14:52:03.282002", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "985d8f45-6e25-480b-9d04-af77b55ba66e", "position": 4, "resourceId": "fab57d92-2a95-4220-ba02-5082a8b6c028", "resource_type": null, "rw_id": "025d67fa-51f0-4051-b7a4-c12ca2209b3a", "schema": {"value": []}, "size": null, "state": "active", "title": "Buffalo headcount in 2022", "type": "reference-layer", "url": "", "url_type": "reference-layer"}, {"cache_last_updated": null, "cache_url": null, "created": "2026-07-13T13:28:57.219000", "datastore_active": false, "description": "Horse headcount for year 2022", "format": "", "hash": "", "id": "b2b5fdac-8a9e-4749-8259-5d9202105b2d", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-07-17T14:52:03.282083", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "985d8f45-6e25-480b-9d04-af77b55ba66e", "position": 5, "resourceId": "b2b5fdac-8a9e-4749-8259-5d9202105b2d", "resource_type": null, "rw_id": "8d322fc9-7529-45b7-8d51-e42e41980268", "schema": {"value": []}, "size": null, "state": "active", "title": "Horse headcount in 2022", "type": "reference-layer", "url": "", "url_type": "reference-layer"}], "tags": [{"display_name": "Agriculture", "id": "d6a3c60e-5e1b-46a7-b418-4cedc8383ca5", "name": "Agriculture", "state": "active", "vocabulary_id": null}, {"display_name": "Grasslands", "id": "f518c799-7d2a-4961-b315-0d4b0e1ac1ab", "name": "Grasslands", "state": "active", "vocabulary_id": null}, {"display_name": "Land Use", "id": "76440e7f-93d0-4b3d-b214-58947224c675", "name": "Land Use", "state": "active", "vocabulary_id": null}, {"display_name": "Livestock", "id": "fe8660f3-8489-480f-b5c4-543db4a46ac1", "name": "Livestock", "state": "active", "vocabulary_id": null}, {"display_name": "Machine learning", "id": "bd190ecb-cb0f-4c96-bfd3-0c0b33d99e91", "name": "Machine learning", "state": "active", "vocabulary_id": null}], "relationships_as_subject": [], "relationships_as_object": []}, {"approval_status": "approved", "author": null, "author_email": null, "authors": [{"name": "Jessica Richter", "email": "Jessica.Richter@wri.org"}, {"name": "Liz Goldman", "email": "elizabeth.goldman@wri.org"}, {"name": "... and 6 others", "email": ""}], "cautions": "This dataset is a compilation of planted tree data from a variety of countries and sources. As a result, there are definitional and temporal inconsistencies within the database, as well as an absence of a uniform accuracy assessment.
This dataset uses harmonized data at national and regional scales. For this reason, it is best used at global, regional, national, state, or provincial scales. This dataset was not created for intended use at hyper-local scales (e.g., tracking individual restoration projects at the municipality or community level).
", "citation": "Richter, J., E. Goldman, N. Harris, D. Gibbs, M. Rose, S. Peyer, S. Richardson, and H. Velappan. 2024. \u201cSpatial Database of Planted Trees (SDPT Version 2.0).\u201d Technical Note. Washington, DC: World Resources Institute. \n\nRichter, J., E. Goldman, N. Harris, D. Gibbs, M. Rose. 2025. \u201cSpatial Database of Planted Trees (SDPT Version 2.1).\u201d Change Log. Washington, DC: World Resources Institute. ", "creator_user_id": "08edbeba-0931-4386-b5d5-63a0a646f718", "draft": false, "featured_dataset": false, "has_chart_views": false, "id": "c93085de-c7fa-4d0a-b3d1-dad65db4cb53", "is_approved": true, "isopen": true, "language": "en", "learn_more": "https://doi.org/10.46830/writn.23.00073", "license_id": "cc-by", "license_title": "Creative Commons Attribution", "license_url": "http://www.opendefinition.org/licenses/cc-by", "maintainer": null, "maintainer_email": null, "maintainers": [{"name": "Jessica Richter", "email": "Jessica.Richter@wri.org"}], "metadata_created": "2026-05-14T15:38:34.271118", "metadata_modified": "2026-06-25T15:30:57.120077", "methodology": "The SDPT was created through cleaning and processing individual planted forest and tree crop datasets, leading to the creation of a harmonized attribute table. This table was then joined to the planted tree vector boundaries and compiled into a vector file geodatabase.
Carbon removal factors and standard deviations were sourced from scientific and gray literature and assigned to unique combinations of planted tree species defined in the harmonized attribute table.
For additional methodological details, please refer to the SDPT version 2.0 technical documentation.
", "name": "spatial-database-of-planted-trees-sdpt-version-21", "notes": "The Spatial Database of Planted Trees (SDPT) was compiled by Global Forest Watch using planted tree boundary data obtained from national governments, non-governmental organizations and independent researchers. SDPT version 2.1 covers 158 countries around the world, with most data sources originating from national level maps, created via supervised classification or manual polygon delineation of Landsat, SPOT or RapidEye satellite imagery. Data source years range from 2000 to 2023, nominally representing 2020. \u201cPlanted trees\u201d in the SDPT includes \u201cplanted forests,\u201d stand of planted trees - other than tree crops - grown for wood and wood fiber production or for ecosystem protection against wind and/or soil erosion, as well as perennial \u201ctree crops,\u201d such as rubber, oil palm, coffee, coconut, cocoa and orchards. The SDPT makes it possible to identify planted forests and tree crops apart from natural forests and enables changes in these planted areas to be monitored independently from changes in global natural forest cover.
Carbon removal factors associated with planted trees were also compiled from a variety of sources. The removal factors, applied spatially to the database based on species information, can be used to map and model carbon removals for planted forests and tree crops, or in comparisons with removal factors by natural forests.
", "num_resources": 2, "num_tags": 5, "open_in": "[{\"title\":\"Global Forest Watch\",\"url\":\"https://www.globalforestwatch.org/map/?map=eyJkYXRhc2V0cyI6W3siZGF0YXNldCI6InRyZWUtcGxhbnRhdGlvbnMiLCJvcGFjaXR5IjoxLCJ2aXNpYmlsaXR5Ijp0cnVlLCJsYXllcnMiOlsidHJlZS1wbGFudGF0aW9ucyJdfSx7ImRhdGFzZXQiOiJwb2xpdGljYWwtYm91bmRhcmllcyIsImxheWVycyI6WyJkaXNwdXRlZC1wb2xpdGljYWwtYm91bmRhcmllcyIsInBvbGl0aWNhbC1ib3VuZGFyaWVzIl0sIm9wYWNpdHkiOjEsInZpc2liaWxpdHkiOnRydWV9XX0%3D\"}]", "organization": {"id": "ad461dbd-f646-4080-94a9-c10005c42a43", "name": "land-carbon-lab", "title": "Land & Carbon Lab", "type": "organization", "description": "WRI's Land & Carbon Lab creates datasets to support critical monitoring and decision making needs surrounding land use planning, carbon accounting, and other technical facets of human-land interactions.", "image_url": "1724353785-cover-ethiopia-mosaic-oqiid6.avif", "created": "2024-09-25T11:07:47.708772", "is_organization": true, "approval_status": "approved", "state": "active"}, "owner_org": "ad461dbd-f646-4080-94a9-c10005c42a43", "private": false, "project": "", "release_notes": "New data sources were added to the Spatial Database of Planted Trees version 2.0 (Richter et al. 2024), prompting the creation of SDPT version 2.1. In this new version, targeted improvements were made to tree crop data sources across four countries: Brazil, Cambodia, Cote d\u2019Ivoire, and Ghana. Supplemental documentation of these improvements can be found in the SDPT version 2.1 change log, available on the SDPT v2 publication.
", "restrictions": "", "rw_dataset": false, "rw_id": "", "short_description": "A database of planted forests and tree crop boundaries at a near-global scale, with associated carbon removal factors.", "spatial": "null", "spatial_address": "Global", "spatial_type": "global", "state": "active", "technical_notes": "https://doi.org/10.46830/writn.23.00073", "title": "Spatial Database of Planted Trees (SDPT Version 2.1)", "type": "dataset", "update_frequency": "as_needed", "version": null, "visibility_type": "public", "wri_data": true, "groups": [{"description": "Data related to tree cover, deforestation, reforestation, and forest biodiversity.", "display_name": "Forests", "id": "c74fae7a-9b54-4c02-b7b5-e3ca00bff8ef", "image_display_url": "", "name": "forests", "title": "Forests", "type": "group"}, {"description": "Data concerning land use, land cover, and terrestrial ecosystem dynamics along with the human and environmental drivers of our food, forests, and water systems.", "display_name": "Land", "id": "c1427a08-9de5-4182-9c05-b6a61fda3127", "image_display_url": "https://datasets.wri.org/private-admin/uploads/group/land-peru-amazon-river-jrjo7f.jpg", "name": "land", "title": "Land", "type": "group"}, {"description": "Global Forest Watch (GFW) is an online platform that provides data and tools for monitoring forests. By harnessing cutting-edge technology, GFW allows anyone to access near real-time information about where and how forests are changing around the world.", "display_name": "Global Forest Watch", "id": "c3429d38-c9ae-443a-8f7e-e20e6415153c", "image_display_url": "https://datasets.wri.org/private-admin/uploads/group/background-268968d448f189afbf5bfc47eb090ac4-5tjpmt.jpeg", "name": "global-forest-watch-app", "title": "Global Forest Watch", "type": "application", "contact_url": "https://www.globalforestwatch.org/about/?contactUs=true", "help_url": "https://www.globalforestwatch.org/help/", "homepage_url": "https://www.globalforestwatch.org/"}], "resources": [{"cache_last_updated": null, "cache_url": null, "created": "2026-05-14T15:38:34.275475", "datastore_active": false, "description": "A vector file geodatabase of planted tree boundaries and associated carbon removal factors.", "format": "ZIP", "hash": "", "id": "6c38393c-a169-4679-a633-9e214aa3c5a0", "last_modified": null, "layer": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-25T15:30:57.131147", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "c93085de-c7fa-4d0a-b3d1-dad65db4cb53", "position": 0, "resourceId": "6c38393c-a169-4679-a633-9e214aa3c5a0", "resource_type": null, "size": null, "state": "active", "title": "Spatial Database of Planted Trees (SDPT version 2.1)", "total_record_count": null, "type": "link", "url": "https://gfw-files.s3.amazonaws.com/plantations/SDPT_v2.1/sdpt_v21_v09152024_public.gdb.zip", "url_type": "link"}, {"cache_last_updated": null, "cache_url": null, "created": "2026-05-28T13:31:10.578000", "datastore_active": false, "description": "", "format": "", "hash": "", "id": "ba80c268-8d73-49e4-b7c3-ad0e42db82fd", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-25T15:30:57.131258", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "c93085de-c7fa-4d0a-b3d1-dad65db4cb53", "position": 1, "resourceId": "ba80c268-8d73-49e4-b7c3-ad0e42db82fd", "resource_type": null, "rw_id": "9581554d-1197-4ed9-b270-de54f664aa41", "schema": {"value": []}, "size": null, "state": "active", "title": "SDPT v2.1 Interactive Map", "type": "reference-layer", "url": "", "url_type": "reference-layer"}], "tags": [{"display_name": "Carbon Sequestration", "id": "3e1eb220-500f-4592-844c-0580cfc3866f", "name": "Carbon Sequestration", "state": "active", "vocabulary_id": null}, {"display_name": "Plantations", "id": "c07b8f21-5f1b-486e-b576-e299c3c5699d", "name": "Plantations", "state": "active", "vocabulary_id": null}, {"display_name": "Planted Forests", "id": "5a040c33-a8b0-4616-8668-f93d2efa10cc", "name": "Planted Forests", "state": "active", "vocabulary_id": null}, {"display_name": "Tree Crop", "id": "e6d6705b-4b21-4c5d-b790-6ddd5c57fd50", "name": "Tree Crop", "state": "active", "vocabulary_id": null}, {"display_name": "Trees", "id": "4d85cc9f-5c79-4576-be65-00689a2cc336", "name": "Trees", "state": "active", "vocabulary_id": null}], "relationships_as_subject": [], "relationships_as_object": []}, {"approval_status": "approved", "author": null, "author_email": null, "authors": [{"name": "Aqueduct Team", "email": ""}], "cautions": "This data is condensed from higher resolution basin-level indicators that are also available.
Key elements of Aqueduct, such as overall water risk, cannot be directly measured and therefore are not validated. Aqueduct remains primarily a prioritization tool and should be augmented by local and regional deep dives.
The aggregation methodology is described in the WRI Technical Note for Aqueduct 4.0 indicators under the section \"Country and State Aggregation\".
", "name": "aqueduct-40-current-and-future-country-rankings", "notes": "Most water-related decisions are made across political or administrative boundaries, creating a demand for simple and robust water information to support decision making at the administrative level. However, accurately assessing the state of water resources across administrative boundaries is a significant challenge; and simple, comparable, and robust water information to support decision making at that level remains sparse.
Aqueduct 4.0 provides catchment-level data on baseline and future water risks but this dataset brings it up to the country and provincial levels using a weighted aggregation methodology based on gross water demand. The gross demand data is used to indicate where the human need for water is greatest\u2014it is also where socioeconomic dependency on water is most critical.
This dataset provides the country and provincial aggregated water risk in tabular form. It includes baseline risks as well as future projections of water risks. The future projections are based on CMIP6 climate projections for 3 scenarios covering 3 milestone years: 2030, 2050, and 2080.
", "num_resources": 1, "num_tags": 7, "open_in": "[{\"title\":\"Aqueduct Country Rankings Tool\",\"url\":\"https://www.wri.org/applications/aqueduct/country-rankings/\"}]", "organization": {"id": "9904135b-b1f3-4fdb-ab86-c25d8d8e99ce", "name": "aqueduct", "title": "Aqueduct", "type": "organization", "description": "WRI's Aqueduct team is part of the Freshwater Program and manages a suite of tools that use open-source, peer reviewed data to map water risks such as floods, droughts and stress. Beyond the tools, the Aqueduct team works one-on-one with companies, governments and research partners through the Aqueduct Alliance to help advance best practices in water resource management and enable sustainable growth in a water-constrained world.", "image_url": "aqueduct-riverine-brazil-osoazh.jpg", "created": "2025-10-27T21:24:26.434769", "is_organization": true, "approval_status": "approved", "state": "active"}, "owner_org": "9904135b-b1f3-4fdb-ab86-c25d8d8e99ce", "private": false, "project": "", "rw_dataset": false, "rw_id": "", "short_description": "Complex hydrological data condensed into intuitive indicators of water stress and further aggregated at the country and provincial levels.", "spatial": "null", "spatial_address": "Global", "spatial_type": "global", "state": "active", "technical_notes": "https://doi.org/10.46830/writn.23.00061", "temporal_coverage_end": "2080", "temporal_coverage_start": "2014", "title": "Aqueduct 4.0 Current and Future Country Rankings", "type": "dataset", "update_frequency": "not_planned", "version": null, "visibility_type": "public", "wri_data": true, "resources": [{"cache_last_updated": null, "cache_url": null, "created": "2026-04-20T17:54:37.428800", "datastore_active": false, "description": "Zipfile (12MB) of MS Excel spreadsheet containing tabular data at country and province level.", "format": "XLS", "hash": "", "id": "589abc06-adae-4d9a-ab69-1a0206a1a608", "last_modified": null, "layer": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-22T13:44:47.534164", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "67e4343d-efa9-4585-a519-b74e12fbfce2", "position": 0, "resourceId": "589abc06-adae-4d9a-ab69-1a0206a1a608", "resource_type": null, "size": null, "state": "active", "title": "Aqueduct 4.0 Current and Future Country Rankings", "total_record_count": null, "type": "link", "url": "https://aqueduct.wridata.org/Aqueduct40/aqueduct-4-0-country-rankings.zip", "url_type": "link"}], "tags": [{"display_name": "Freshwater", "id": "fbfdb860-3f07-4524-b9eb-0b60c3e3c808", "name": "Freshwater", "state": "active", "vocabulary_id": null}, {"display_name": "Groundwater", "id": "9216b881-d157-42be-aa23-c24602fc2fdb", "name": "Groundwater", "state": "active", "vocabulary_id": null}, {"display_name": "Hydrology", "id": "17e1ad89-4a77-4405-8abd-6fa4fff1fcc1", "name": "Hydrology", "state": "active", "vocabulary_id": null}, {"display_name": "Water Availability", "id": "fa0dd2e6-38a4-49ea-9343-90009abaf5c0", "name": "Water Availability", "state": "active", "vocabulary_id": null}, {"display_name": "Water Quality", "id": "f68c7ff5-c19c-4239-b607-60645c503f7c", "name": "Water Quality", "state": "active", "vocabulary_id": null}, {"display_name": "Water Risk", "id": "81f2e442-d368-49a6-9f27-ddc2fdbdd66a", "name": "Water Risk", "state": "active", "vocabulary_id": null}, {"display_name": "Water Stress", "id": "c739c8c8-1431-44a2-8dc8-eb083408ab2b", "name": "Water Stress", "state": "active", "vocabulary_id": null}], "groups": [], "relationships_as_subject": [], "relationships_as_object": []}, {"approval_status": "approved", "author": null, "author_email": null, "authors": [{"name": "Sam Kuzma", "email": "samantha.kuzma@wri.org"}], "cautions": "This data was released in 2020 and reflects research and modeling capabilities at that time.
The hazard layers have been simulated without considering the presence of flood protection.
Section 3 of the Technical Note describes the flood hazard modeling and it is fully documented in Appendix A.1.
Briefly:
Riverine flooding:
Performed with GLOFRIS and PCR-GLOBWB models forced by EUWATCH and IS-MIP climate meteorological datasets.
Model period configured to 1950 - 2099; 1960-1999 is the baseline/historic period.
Model outputs at 5' x 5' refined to 30\" x 30\".
Coastal flooding:
Start from GTSR as database of extreme water levels for baseline/historic flood inundation depths.
Future projections based on gridded sea-level changes from the RISES-AM project.
Subsidence is estimated by an ensemble of models.
Aqueduct Floods files are distributed as single-band 32-bit float Geotiffs with pixels corresponding the inundation depth in meters for a specific flood return period. The value of -9999 is used as NoData.
There are a high number of model parameterizations and measurements enabling users to compare different scenarios for riverine and coastal flooding. Common to both the coastal and riverine flood modeling are:
Climate scenarios
Historical baseline scenario
Optimistic future scenario corresponding to RCP 4.5
Pessimistic future scenario corresponding to RCP 8.5
Timeslices of output
Current/historic for baseline model
2030
2050
2080
Riverine floods are further parameterized by the hydrologic model employed
NorESM1-M
GFDL-ESM2M
HadGEM2-ES
IPSL-CM5A-LR
MIROC-ESM-CHEM
Access to the files and supporting documentation is provided through a standalone webpage.
", "num_resources": 2, "num_tags": 6, "open_in": "[{\"title\":\"Aqueduct Floods\",\"url\":\"https://www.wri.org/applications/aqueduct/floods/\"}]", "organization": {"id": "9904135b-b1f3-4fdb-ab86-c25d8d8e99ce", "name": "aqueduct", "title": "Aqueduct", "type": "organization", "description": "WRI's Aqueduct team is part of the Freshwater Program and manages a suite of tools that use open-source, peer reviewed data to map water risks such as floods, droughts and stress. Beyond the tools, the Aqueduct team works one-on-one with companies, governments and research partners through the Aqueduct Alliance to help advance best practices in water resource management and enable sustainable growth in a water-constrained world.", "image_url": "aqueduct-riverine-brazil-osoazh.jpg", "created": "2025-10-27T21:24:26.434769", "is_organization": true, "approval_status": "approved", "state": "active"}, "owner_org": "9904135b-b1f3-4fdb-ab86-c25d8d8e99ce", "private": false, "project": "Aqueduct Floods", "rw_dataset": false, "rw_id": "", "short_description": "Flood inundation depth maps that support the Aqueduct Floods online tool. This dataset provides estimations of inundation depth in meters for riverine and coastal floods under both current baseline conditions and future projections in 2030, 2050, and 2080 at several levels of flood intensity return periods (recurrence intervals).", "spatial": "null", "spatial_address": "Global", "spatial_type": "global", "state": "active", "technical_notes": "https://www.wri.org/research/aqueduct-floods-methodology", "temporal_coverage_end": "2080", "temporal_coverage_start": "2010", "title": "Aqueduct Floods Hazard Maps", "type": "dataset", "update_frequency": "not_planned", "url": "https://www.wri.org/data/aqueduct-floods-hazard-maps", "version": null, "visibility_type": "public", "wri_data": true, "resources": [{"cache_last_updated": null, "cache_url": null, "created": "2025-11-03T16:06:30.966738", "datastore_active": false, "description": "The downloadable hazard maps were updated on October 20, 2020 to fix a bias-correction error in the data. The update only impacts the riverine maps. If you downloaded hazard maps prior to October 20, 2020, we recommend you redownload the data.", "format": "", "hash": "", "id": "9d15d12a-e813-49ae-a8fd-9cbc71c36138", "last_modified": null, "layer": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-22T13:41:30.673380", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": true, "package_id": "7cbe47a3-e256-47d5-ad00-e2ddc9d7f9d9", "position": 0, "resourceId": "9d15d12a-e813-49ae-a8fd-9cbc71c36138", "resource_type": null, "size": null, "state": "active", "title": "Aqueduct Floods Hazard Maps, Version 2.0", "total_record_count": null, "type": "link", "url": "https://aqueduct.wridata.org/AqueductFloods20/index.html", "url_type": "link"}, {"cache_last_updated": null, "cache_url": null, "created": "2026-04-07T19:22:51.753000", "datastore_active": false, "description": "Riverine flood hazard from GFDL-ESM2M model; Year 2050; Business as Usual scenario; 250 year return period", "format": "", "hash": "", "id": "113c974f-669e-4dbb-97af-fd8fe4657894", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-22T13:41:30.673489", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "7cbe47a3-e256-47d5-ad00-e2ddc9d7f9d9", "position": 1, "resourceId": "113c974f-669e-4dbb-97af-fd8fe4657894", "resource_type": null, "rw_id": "d9845d4a-8313-460f-ac7d-1c330da905a6", "schema": {"value": []}, "size": null, "state": "active", "title": "Riverine flood map layer", "type": "reference-layer", "url": "", "url_type": "reference-layer"}], "tags": [{"display_name": "Coastal Flooding", "id": "6b053131-c8f3-40c0-871b-a49dcafdd324", "name": "Coastal Flooding", "state": "active", "vocabulary_id": null}, {"display_name": "Flood Risk", "id": "93cd8636-002c-43bc-9f37-d1cf4d37e85d", "name": "Flood Risk", "state": "active", "vocabulary_id": null}, {"display_name": "Floods", "id": "d25ebee7-33d9-4936-a729-83515f9dfc82", "name": "Floods", "state": "active", "vocabulary_id": null}, {"display_name": "Hydrology", "id": "17e1ad89-4a77-4405-8abd-6fa4fff1fcc1", "name": "Hydrology", "state": "active", "vocabulary_id": null}, {"display_name": "Riverine Flooding", "id": "9ab27ded-b943-47bb-a0df-747096d1d2c6", "name": "Riverine Flooding", "state": "active", "vocabulary_id": null}, {"display_name": "Surface Water", "id": "45627f0f-577f-442e-87f2-03bbdbb6fdf2", "name": "Surface Water", "state": "active", "vocabulary_id": null}], "groups": [], "relationships_as_subject": [], "relationships_as_object": []}, {"approval_status": "approved", "author": null, "author_email": null, "authors": [{"name": "Leah Lazer", "email": null}, {"name": "Lydia Freehafer", "email": null}, {"name": "Brian Zepka", "email": "brian.zepka@wri.org"}], "cautions": "This dataset is the result of new data collection by WRI\u2019s Electric School Bus Initiative, and is sourced from hundreds of news articles, school district webpages, and other online sources.\nTo the best of our knowledge, these data are up to date as of the time described in documentation packaged with each version.\nThe data represent a snapshot in time, in a rapidly evolving space.
", "citation": "Lazer, L., L. Freehafer, and B. Zepka. June 2025. Dataset of Electric School Bus Adoption in the United States. Version 9. World Resources Institute. Accessed through https://datasets.wri.org/dataset/electric-school-bus-adoption onESB-related data were collected from a variety of publicly available sources, including news articles, school websites, industry publications like School Bus Fleet magazine, and social media posts.\nOther demographic and economic data come from reputable, public datasets including the Environmental Protection Agency (EPA), U.S. Census, and National Center for Education Statistics.
", "name": "electric-school-bus-adoption", "notes": "Transitioning to electric school buses (ESBs) from traditional diesel-powered school buses can reduce students\u2019 exposure to air pollution and decrease greenhouse gas emissions.\nSchool districts and private fleet operators around the United States are adopting electric school buses with increasing speed, but so far ESB adoption has not been tracked or reported in a centralized and publicly accessible way.\nWRI aims to create and disseminate sound, up-to-date, accessible data and analyses that can help school district staff, advocates, policymakers, and other stakeholders make evidence-based decisions and support the transition to electric school buses.\nThis first-of-its-kind dataset that tracks ESB adoption across the United States.
\nThe dataset is organized by both school district and individual ESB and tracks the number of \"committed\" ESBs.\nAn ESB is considered \"committed\" starting from the point when a school district or fleet operator has been awarded funding to purchase it or has made formal agreement to purchase it from a manufacturer or dealer.\nWe would not consider an ESB \"committed\" if a school district or other fleet operator only expressed interest in ESBs or stated that they plan to acquire ESBs, without awarded funding or an agreement with a third party.\nThe dataset also tracks the progress of each individual ESB through the four phases of the adoption process: \"awarded\", \"ordered\", \"delivered\", and \"operating\".\nIt also contains school district characteristics including poverty, racial composition, air pollution, and locale (urban, suburban, town, or rural), to enable wider analysis of the adoption of ESBs, including the extent to which the transition to ESBs is happening equitably.
", "num_resources": 10, "num_tags": 2, "open_in": "[{\"title\":\"Electric School Bus Initiative Dashboard\",\"url\":\"https://electricschoolbusinitiative.org/electric-school-bus-data-dashboard\"}]", "organization": {"id": "164c340b-5cf2-49da-98f9-024fdb7e0e42", "name": "electric-school-bus-initiative", "title": "Electric School Bus Initiative", "type": "organization", "description": "WRI's Electric School Bus Initiative accelerates the adoption of electric school buses to deliver on improved health and environmental outcomes in communities across the United States of America.", "image_url": "students-boarding-school-bus-m3277f.jpg", "created": "2024-09-24T18:14:31.852003", "is_organization": true, "approval_status": "approved", "state": "active"}, "owner_org": "164c340b-5cf2-49da-98f9-024fdb7e0e42", "private": false, "project": "", "rw_dataset": false, "rw_id": "", "short_description": "This dataset tracks electric school bus (ESB) adoption across the United States. It tracks the number of \"committed\" ESBs at the school district level, as well as details about individual buses, including the bus manufacturer and funding source(s). It also tracks when each ESB passed through the phases of the adoption process and the current phase of each bus. The dataset contains school district socio-economic characteristics, like poverty rates, racial composition and air pollution to enable wider analysis including whether the transition to ESBs is happening equitably.", "spatial": "null", "spatial_address": "United States", "spatial_type": "address", "state": "active", "technical_notes": "https://www.wri.org/research/technical-note-dataset-electric-school-bus-adoption-united-states", "title": "Dataset of Electric School Bus Adoption in the United States", "type": "dataset", "update_frequency": "biannually", "version": null, "visibility_type": "public", "wri_data": true, "groups": [{"description": "Data concerning urban spaces and systems.", "display_name": "Cities", "id": "80539133-5aaa-4257-9aeb-fe6f56e2837f", "image_display_url": "https://datasets.wri.org/private-admin/uploads/group/children-biking-calle-colombia-2-qy6gqa.jpeg", "name": "cities", "title": "Cities", "type": "group"}, {"description": "Data concerning transportation systems including public and private transit.", "display_name": "Mobility", "id": "4926571e-a4e5-403d-8278-31a570039cc3", "image_display_url": "", "name": "mobility", "title": "Mobility", "type": "group"}], "resources": [{"cache_last_updated": null, "cache_url": null, "created": "2025-06-23T18:14:44.432397", "datastore_active": false, "description": "Released 2022-January (data through 2021-December)", "format": "ZIP", "hash": "", "id": "703943f3-a051-4829-a65d-58eef9002479", "last_modified": "2025-06-23T18:14:44.291027", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.778698", "mimetype": "application/zip", "mimetype_inner": null, "name": "Electric School Bus Adoption Version 1", "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 0, "resourceId": "703943f3-a051-4829-a65d-58eef9002479", "resource_type": null, "size": 9116566, "state": "active", "title": "Electric School Bus Adoption Version 1", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/917ef3ec-0f73-4188-b7c1-31bb7a375a4c/resource/703943f3-a051-4829-a65d-58eef9002479/download/electric_school_bus_adoption_dataset_v1_2022-01jan.zip", "url_type": "upload"}, {"cache_last_updated": null, "cache_url": null, "created": "2025-06-23T18:14:46.107616", "datastore_active": false, "description": "Released 2022-May (data through 2022-March)", "format": "ZIP", "hash": "", "id": "038ff462-0072-4fcc-a7a1-264a885750c8", "last_modified": "2025-06-23T18:14:45.821146", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.778807", "mimetype": "application/zip", "mimetype_inner": null, "name": "Electric School Bus Adoption Version 2", "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 1, "resourceId": "038ff462-0072-4fcc-a7a1-264a885750c8", "resource_type": null, "size": 15317571, "state": "active", "title": "Electric School Bus Adoption Version 2", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/917ef3ec-0f73-4188-b7c1-31bb7a375a4c/resource/038ff462-0072-4fcc-a7a1-264a885750c8/download/electric_school_bus_adoption_dataset_v2_2022-05may.zip", "url_type": "upload"}, {"cache_last_updated": null, "cache_url": null, "created": "2025-06-23T18:14:48.337640", "datastore_active": false, "description": "Released 2022-August (data through 2022-June)", "format": "ZIP", "hash": "", "id": "fc18678f-e3c1-4a7b-aad4-d01f65f92681", "last_modified": "2025-06-23T18:14:47.945362", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.778886", "mimetype": "application/zip", "mimetype_inner": null, "name": "Electric School Bus Adoption Version 3", "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 2, "resourceId": "fc18678f-e3c1-4a7b-aad4-d01f65f92681", "resource_type": null, "size": 17296237, "state": "active", "title": "Electric School Bus Adoption Version 3", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/917ef3ec-0f73-4188-b7c1-31bb7a375a4c/resource/fc18678f-e3c1-4a7b-aad4-d01f65f92681/download/electric_school_bus_adoption_dataset_v3_2022-08aug.zip", "url_type": "upload"}, {"cache_last_updated": null, "cache_url": null, "created": "2025-06-23T18:14:50.913200", "datastore_active": false, "description": "Released 2023-February (data through 2022-September)", "format": "ZIP", "hash": "", "id": "882b23b9-afec-4fa3-a070-8b91463872d5", "last_modified": "2025-06-23T18:14:50.439541", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.778963", "mimetype": "application/zip", "mimetype_inner": null, "name": "Electric School Bus Adoption Version 4", "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 3, "resourceId": "882b23b9-afec-4fa3-a070-8b91463872d5", "resource_type": null, "size": 18617838, "state": "active", "title": "Electric School Bus Adoption Version 4", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/917ef3ec-0f73-4188-b7c1-31bb7a375a4c/resource/882b23b9-afec-4fa3-a070-8b91463872d5/download/electric_school_bus_adoption_dataset_v4_2023-02feb.zip", "url_type": "upload"}, {"cache_last_updated": null, "cache_url": null, "created": "2025-06-23T18:14:53.728004", "datastore_active": false, "description": "Released 2023-March (data through 2022-December)", "format": "ZIP", "hash": "", "id": "48c9d3b1-084b-4de8-ab24-8e444f8c61cc", "last_modified": "2025-06-23T18:14:53.127406", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.779034", "mimetype": "application/zip", "mimetype_inner": null, "name": "Electric School Bus Adoption Version 5", "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 4, "resourceId": "48c9d3b1-084b-4de8-ab24-8e444f8c61cc", "resource_type": null, "size": 18977181, "state": "active", "title": "Electric School Bus Adoption Version 5", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/917ef3ec-0f73-4188-b7c1-31bb7a375a4c/resource/48c9d3b1-084b-4de8-ab24-8e444f8c61cc/download/electric_school_bus_adoption_dataset_v5_2023-03mar.zip", "url_type": "upload"}, {"cache_last_updated": null, "cache_url": null, "created": "2025-06-23T18:14:56.295544", "datastore_active": false, "description": "Released 2023-August (data through 2023-June)", "format": "ZIP", "hash": "", "id": "caf199a4-ba83-438e-a706-c380d5de15b4", "last_modified": "2025-06-23T18:14:55.570146", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.779109", "mimetype": "application/zip", "mimetype_inner": null, "name": "Electric School Bus Adoption Version 6", "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 5, "resourceId": "caf199a4-ba83-438e-a706-c380d5de15b4", "resource_type": null, "size": 19491886, "state": "active", "title": "Electric School Bus Adoption Version 6", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/917ef3ec-0f73-4188-b7c1-31bb7a375a4c/resource/caf199a4-ba83-438e-a706-c380d5de15b4/download/electric_school_bus_adoption_dataset_v6_2023-08aug.zip", "url_type": "upload"}, {"cache_last_updated": null, "cache_url": null, "created": "2025-06-23T18:14:59.220081", "datastore_active": false, "description": "Released 2024-April (data through 2023-December)", "format": "ZIP", "hash": "", "id": "ddb521da-32ee-4aef-8dc4-6a807017ea35", "last_modified": "2025-06-23T18:14:58.410789", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.779179", "mimetype": "application/zip", "mimetype_inner": null, "name": "Electric School Bus Adoption Version 7", "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 6, "resourceId": "ddb521da-32ee-4aef-8dc4-6a807017ea35", "resource_type": null, "size": 19637829, "state": "active", "title": "Electric School Bus Adoption Version 7", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/917ef3ec-0f73-4188-b7c1-31bb7a375a4c/resource/ddb521da-32ee-4aef-8dc4-6a807017ea35/download/electric_school_bus_adoption_dataset_v7_2024-04apr.zip", "url_type": "upload"}, {"cache_last_updated": null, "cache_url": null, "created": "2025-06-23T18:15:02.778456", "datastore_active": false, "description": "Released 2024-August (data through 2024-June)", "format": "ZIP", "hash": "", "id": "f79d6697-e6e7-4c16-af9a-69503e716ffe", "last_modified": "2025-06-23T18:15:01.785144", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.779247", "mimetype": "application/zip", "mimetype_inner": null, "name": "Electric School Bus Adoption Version 8", "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 7, "resourceId": "f79d6697-e6e7-4c16-af9a-69503e716ffe", "resource_type": null, "size": 18628822, "state": "active", "title": "Electric School Bus Adoption Version 8", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/917ef3ec-0f73-4188-b7c1-31bb7a375a4c/resource/f79d6697-e6e7-4c16-af9a-69503e716ffe/download/electric_school_bus_adoption_dataset_v8_2024-08aug.zip", "url_type": "upload"}, {"cache_last_updated": null, "cache_url": null, "created": "2025-06-23T18:15:05.755751", "datastore_active": false, "description": "Released 2025-June (data through 2024-December)", "format": "ZIP", "hash": "", "id": "7fe159da-9a5f-4d75-b02e-a3316f48291e", "last_modified": "2025-06-23T18:15:04.757945", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.779331", "mimetype": "application/zip", "mimetype_inner": null, "name": "Electric School Bus Adoption Version 9", "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 8, "resourceId": "7fe159da-9a5f-4d75-b02e-a3316f48291e", "resource_type": null, "size": 16617301, "state": "active", "title": "Electric School Bus Adoption Version 9", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/917ef3ec-0f73-4188-b7c1-31bb7a375a4c/resource/7fe159da-9a5f-4d75-b02e-a3316f48291e/download/electric_school_bus_adoption_dataset_v9_2025-06jun.zip", "url_type": "upload"}, {"cache_last_updated": null, "cache_url": null, "created": "2026-06-04T17:24:03.057000", "datastore_active": false, "description": "Released 2026-June (data through 2025-December)", "format": "ZIP", "hash": "", "id": "f7d239be-047e-41d8-8cbb-a66a9b237b17", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-06-04T17:24:47.780864", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "917ef3ec-0f73-4188-b7c1-31bb7a375a4c", "position": 9, "resourceId": "f7d239be-047e-41d8-8cbb-a66a9b237b17", "resource_type": null, "schema": {"value": []}, "size": null, "state": "active", "title": "Electric School Bus Adoption Version 10", "type": "link", "url": "https://wri-datalab-public-files.s3.us-east-1.amazonaws.com/datasets/esbi/esb-adoption/electric_school_bus_adoption_dataset_v10_2026-06jun.zip", "url_type": "link"}], "tags": [{"display_name": "education", "id": "98308223-4dfa-43b1-8de6-742cc5c54997", "name": "education", "state": "active", "vocabulary_id": null}, {"display_name": "electric vehicles", "id": "2e6e58ab-8c21-4f4f-ba68-48c6a5183efd", "name": "electric vehicles", "state": "active", "vocabulary_id": null}], "relationships_as_subject": [], "relationships_as_object": []}, {"approval_status": "approved", "author": null, "author_email": null, "authors": [{"name": "John Brandt", "email": "john.brandt@wri.org"}, {"name": "Jessica Ertel", "email": "jessica.ertel@wri.org"}, {"name": "Justine Spore", "email": "justine.spore@wri.org"}, {"name": "Fred Stolle", "email": "fred.stolle@wri.org"}], "cautions": "This dataset uses a different definition of a tree and a different definition of tree cover than does Hansen et al. (2013).
This dataset defines a tree according to both the height and crown diameter. Woody vegetation higher than 5 meters regardless of crown diameter, or between 3 and 5 meters with a minimum crown diameter of 5 meters is considered a tree. This definition is different from Hansen et al. (2013) which defines a tree as any vegetation at least 5 meters in height.
The tropical tree cover dataset does not disambiguate plantation trees from non-plantation trees.
Analyses or statistics derived over spatial regions smaller than 0.5 ha may not be accurate.
Please review the product specification documentation for comprehensive descriptions of the data creation and validation methods.
This dataset utilizes Sentinel-2 and Sentinel-1 optical and radar satellite data with a time-series convolutional neural network to perform image segmentation on monthly composite images.
Some examples of preprocessing and data processing choices:
Capping the maximum number of images per year at 12 to reduce data throughput requirements
Improving Sentinel-2 image selection by instituting inter-quartile range (IQR) thresholds
Instituting saturation thresholds and azimuth/zenith thresholds to improve the quality of input images
Applying terrain flattening to the Sentinel-1 data
Implementing improved cloud shadow detection
Model training and validation data were labeled by the authors. A total of 18,100 sample plots sized 140 x 140 meters were labeled with sampling points positioned at 10 meter intervals for 196 samples per plot. Pixels were marked positive if their centroid intersected a tree, as identified through photointerpretation of high-resolution satellite imagery.
", "name": "tropical-tree-cover", "notes": "The tropical tree cover data maps tree extent at the ten-meter scale and tree cover at the half hectare scale to enable accurate monitoring of trees in urban areas, agricultural lands, and in open canopy and dry forest ecosystems. The data extends over 4.3 billion hectares of the global tropics.
The data is derived from multi-temporal convolutional neural network models applied to Sentinel optical and radar imagery. The 10-meter dataset is a binary tree extent layer that is similar to a land cover map, while the tree cover data represents fractional cover at a half-hectare scale.
More details on the methodology and analyses can be found on the GitHub page which holds code and methodologies.
", "num_resources": 2, "num_tags": 6, "open_in": "[{\"title\":\"Global Forest Watch\",\"url\":\"https://www.globalforestwatch.org/map/?ap3c=IGbZ8wQsN-PH3wQEAGbZ8wQxQ-trL60p9nh6hKka3aSnFBgWBA&map=eyJkYXRhc2V0cyI6W3siZGF0YXNldCI6InRyb3BpY2FsLXRyZWUtY292ZXIiLCJvcGFjaXR5IjoxLCJ2aXNpYmlsaXR5Ijp0cnVlLCJsYXllcnMiOlsidHJvcGljYWwtdHJlZS1jb3Zlci1tZXRlcnMiXX0seyJkYXRhc2V0IjoicG9saXRpY2FsLWJvdW5kYXJpZXMiLCJsYXllcnMiOlsiZGlzcHV0ZWQtcG9saXRpY2FsLWJvdW5kYXJpZXMiLCJwb2xpdGljYWwtYm91bmRhcmllcyJdLCJvcGFjaXR5IjoxLCJ2aXNpYmlsaXR5Ijp0cnVlfV19&mapMenu=eyJtZW51U2VjdGlvbiI6ImRhdGFzZXRzIiwiZGF0YXNldENhdGVnb3J5IjoibGFuZENvdmVyIn0%3D\"}]", "organization": {"id": "ad461dbd-f646-4080-94a9-c10005c42a43", "name": "land-carbon-lab", "title": "Land & Carbon Lab", "type": "organization", "description": "WRI's Land & Carbon Lab creates datasets to support critical monitoring and decision making needs surrounding land use planning, carbon accounting, and other technical facets of human-land interactions.", "image_url": "1724353785-cover-ethiopia-mosaic-oqiid6.avif", "created": "2024-09-25T11:07:47.708772", "is_organization": true, "approval_status": "approved", "state": "active"}, "owner_org": "ad461dbd-f646-4080-94a9-c10005c42a43", "private": false, "project": "", "rw_dataset": false, "rw_id": "", "short_description": "This layer displays tree extent at the ten-meter scale and tree cover at the half hectare scale to enable accurate monitoring of trees in urban areas, agricultural lands, and in open canopy and dry forest ecosystems.", "spatial": "null", "spatial_type": "derived_from_resources", "state": "active", "technical_notes": "https://doi.org/10.1016/j.rse.2023.113574", "temporal_coverage_end": "2020", "temporal_coverage_start": "2020", "title": "Tropical Tree Cover", "type": "dataset", "update_frequency": "as_needed", "version": null, "visibility_type": "public", "wri_data": true, "groups": [{"description": "Data concerning land use, land cover, and terrestrial ecosystem dynamics along with the human and environmental drivers of our food, forests, and water systems.", "display_name": "Land", "id": "c1427a08-9de5-4182-9c05-b6a61fda3127", "image_display_url": "https://datasets.wri.org/private-admin/uploads/group/land-peru-amazon-river-jrjo7f.jpg", "name": "land", "title": "Land", "type": "group"}, {"description": "Global Forest Watch (GFW) is an online platform that provides data and tools for monitoring forests. By harnessing cutting-edge technology, GFW allows anyone to access near real-time information about where and how forests are changing around the world.", "display_name": "Global Forest Watch", "id": "c3429d38-c9ae-443a-8f7e-e20e6415153c", "image_display_url": "https://datasets.wri.org/private-admin/uploads/group/background-268968d448f189afbf5bfc47eb090ac4-5tjpmt.jpeg", "name": "global-forest-watch-app", "title": "Global Forest Watch", "type": "application", "contact_url": "https://www.globalforestwatch.org/about/?contactUs=true", "help_url": "https://www.globalforestwatch.org/help/", "homepage_url": "https://www.globalforestwatch.org/"}], "resources": [{"cache_last_updated": null, "cache_url": null, "created": "2026-05-21T13:44:38.754633", "datastore_active": false, "description": "Tropical Tree Cover at 40% probability threshold, indicating where there is an estimated 40% or greater probability of a tree occurring within a 10-meter pixel.", "format": "", "hash": "", "id": "bd567d54-c4b5-4cff-bfc4-417ea5b8ea9b", "last_modified": null, "layer": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-05-21T15:15:24.430172", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "74b54aef-0846-4f5c-ba27-48ba09d44891", "position": 0, "resourceId": "bd567d54-c4b5-4cff-bfc4-417ea5b8ea9b", "resource_type": null, "rw_id": "2431d806-78bc-4453-8963-7200a4089108", "size": null, "state": "active", "title": "Tropical Tree Cover (40% threshold)", "total_record_count": null, "type": "reference-layer", "url": "", "url_type": "reference-layer"}, {"cache_last_updated": null, "cache_url": null, "created": "2026-05-21T13:47:34.211000", "data_api_asset_id": "b08026ff-3280-4a91-b92a-7ff6382e9e50", "data_api_dataset_id": "wri_tropical_tree_cover_extent", "data_api_tiles": ["/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_000E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_100W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_040E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_110E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_000E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30S_010E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30S_030E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_160E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_050W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_060E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_130E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_030E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_080W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_120E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_020W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_050E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_150E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_020W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_100W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_070E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_010E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_110W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30S_020E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_090W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_110W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_160E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_110E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_090W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_040W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_010W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_100W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_150E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_040E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_090E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_040E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_110E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_090W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_040W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_070W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_080E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_160E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_080W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_120E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_080E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_040E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_150E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_070W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_060W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_070E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_130E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_100W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_060W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_010E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_090E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_050W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_080W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_040E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_020W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_140E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_170E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_120E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_060W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_120E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_080E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_110E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_100E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_110E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_000E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_080W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_140E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_030E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_140E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_040E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_020E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_090E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_130E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_100E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/30N_100E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_070E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_070W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_090E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_010E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_000E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_120E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_080W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_010E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_010W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_030E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_090W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_020E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_010W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20N_100E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_060W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_030E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_080W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/20S_020E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_010E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_020E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_050W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_070W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_050W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_070W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_020E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/00N_060W.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10N_030E.tif", "/vsis3/gfw-data-lake/wri_tropical_tree_cover_extent/v20220922/raster/epsg-4326/10/100000/decile/geotiff/10S_030E.tif"], "data_api_version": "v20220922", "datastore_active": false, "description": "Probability of one or more tree canopies intersecting the center of 10-meter spatial resolution pixel. Probabilities are defined at decile levels. Tiles are 10-deg x 10-deg in extent and named after the upper-left (northeast) corner location.", "format": "", "hash": "", "id": "c5e5a16d-9e93-495d-9ff1-007b9d5c40be", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-05-21T15:15:24.430288", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "74b54aef-0846-4f5c-ba27-48ba09d44891", "position": 1, "resourceId": "c5e5a16d-9e93-495d-9ff1-007b9d5c40be", "resource_type": null, "schema": {"value": []}, "size": null, "state": "active", "title": "GeoTIFFs", "type": "data-api-dataset", "url": "", "url_type": "data-api-dataset"}], "tags": [{"display_name": "Forest Cover", "id": "a4228c83-2d3d-4df6-bc67-65ed6f5cb9a8", "name": "Forest Cover", "state": "active", "vocabulary_id": null}, {"display_name": "Forest and Landscape Restoration", "id": "95225628-e5a8-4ddb-96b4-02ef76e6879c", "name": "Forest and Landscape Restoration", "state": "active", "vocabulary_id": null}, {"display_name": "Land Cover", "id": "929b5fb0-876f-42f9-98f4-eb9e98529167", "name": "Land Cover", "state": "active", "vocabulary_id": null}, {"display_name": "Tree Cover", "id": "b888e8c1-a632-4f27-9ff4-074602eefc5c", "name": "Tree Cover", "state": "active", "vocabulary_id": null}, {"display_name": "Trees", "id": "4d85cc9f-5c79-4576-be65-00689a2cc336", "name": "Trees", "state": "active", "vocabulary_id": null}, {"display_name": "Tropical Forests", "id": "49823ce7-c41f-4998-a4d8-394203da849c", "name": "Tropical Forests", "state": "active", "vocabulary_id": null}], "relationships_as_subject": [], "relationships_as_object": []}, {"approval_status": "approved", "author": null, "author_email": null, "authors": [{"name": "Michelle Sims", "email": "michelle.sims@wri.org"}, {"name": "Radost Stanimirova", "email": "radost.stanimirova@wri.org"}], "cautions": "This product shows the dominant driver in each 1 km cell over the entire period. It does not show multiple drivers if they occur in the same cell at smaller scales, nor does it detail the sequence of drivers if multiple occurred at different times within the period.
This product does not distinguish between the loss of natural forest and planted trees (e.g., plantations, tree crops, or agroforestry systems). While tree cover loss associated with the permanent agriculture, hard commodities, and settlements & infrastructure classes represent a close approximation of deforestation, they do not always represent the conversion of natural forests to other land uses and in some cases may represent loss of planted trees. Similarly, replacement of natural forest with wood fiber plantations is not distinguished from routine harvesting within existing plantations established before 2000, as these are both included in the logging class.
These data are limited in scope to attributing drivers to tree cover loss as mapped by the Hansen et al. (2013) tree cover loss product, and therefore the detection of loss is subject to the accuracy of that product. A full description of limitations is included in the publication.
Global map of the dominant driver of tree cover loss at 0.01\u00b0 resolution (~1km) for the period 2001-2025. This is the latest update to this dataset.
", "has_chart_views": false, "id": "3c1f6c9d-c6c3-4b77-9303-25d0b9c171f9", "is_approved": true, "isopen": true, "language": "en", "learn_more": "https://www.globalforestwatch.org/blog/data-and-tools/new-drivers-data-forest-loss", "license_id": "cc-by", "license_title": "Creative Commons Attribution", "license_url": "http://www.opendefinition.org/licenses/cc-by", "maintainer": null, "maintainer_email": null, "maintainers": [{"name": "Michelle Sims", "email": "michelle.sims@wri.org"}, {"name": "Radost Stanimirova", "email": "radost.stanimirova@wri.org"}], "metadata_created": "2025-06-17T13:18:28.142249", "metadata_modified": "2026-05-05T16:35:57.739846", "methodology": "These data were produced by the World Resources Institute and Google DeepMind. The data were developed using a global neural network model (ResNet) trained on a set of 6,955 samples collected through visual interpretation of very high-resolution satellite imagery. The model used satellite imagery (Landsat 7 & 8, Sentinel-2) and ancillary data to classify the seven driver categories.
Overall accuracy of the model is 90.5%, with regional accuracies varying from 82.8% in Southeast Asia to 94.1% in Asia. Global per class producer\u2019s and user\u2019s accuracy are highest for the permanent agriculture, logging, and wildfire classes (over 90%), and generally lower for rarer classes, such as hard commodities, settlements and infrastructure, and other natural disturbances. A full description of the methods and accuracy statistics are available in the publication.
", "name": "dominant-drivers-of-tree-cover-loss-at-1km", "notes": "This dataset contains the dominant driver of tree cover loss from 2001-2025. A driver is defined as the direct cause of tree cover loss, and can include both temporary disturbances (natural or anthropogenic) or permanent loss of tree cover due to a change to a non-forest land use (e.g., deforestation). The dominant driver is defined as the direct driver that caused the majority of tree cover loss within each 1 km cell over a time period.
Classes of drivers are defined as follows:
Permanent agriculture: Long-term, permanent tree cover loss for small- to large-scale agriculture.
Hard commodities: Loss due to the establishment or expansion of mining or energy infrastructure.
Shifting cultivation: Tree cover loss due to small- to medium-scale clearing for temporary cultivation that is later abandoned and followed by subsequent regrowth of secondary forest or vegetation.
Logging: Forest management and logging activities occurring within managed, natural or semi-natural forests and plantations, often with evidence of forest regrowth or planting in subsequent years.
Wildfire: Tree cover loss due to fire with no visible human conversion or agricultural activity afterward. Fires may be started by natural causes (e.g. lightning) or may be related to human activities (accidental or deliberate).
Settlements and infrastructure: Tree cover loss due to expansion and intensification of roads, settlements, urban areas, or built infrastructure (not associated with other classes).
Other natural disturbances: Tree cover loss due to other non-fire natural disturbances (e.g., landslides, insect outbreaks, river meandering). If loss due to natural causes is followed by salvage or sanitation logging, it is classified as logging.
The data is available for download as a global raster file with 8 bands, 8-bit pixel values, and 0.01-degree spatial resolution. The first band is a classification of the dominant driver for the full time period (2001-2025). The pixel values 1 through 7 correspond to the seven classifications of drivers and 255 corresponds to NoData. The other seven bands contain the probability of each driver. Probabilities should be interpreted as a relative confidence between the classes, rather than a true estimate of the likelihood of each class.
The pixel values for the probability bands should NOT be interpreted as an integer percentage (i.e. 1 means 1%, 2 means 2%, ... is false). Instead more dynamic range is provided and the percentage contribution between 0%-100% is quantized into 250 levels each corresponding to a 0.4% increment rather than a 1% increment. The formula that can be used to calculate the percentage probability for a driver is driver_probability_pct = uint8(pixel_value) * 0.4 . Valid pixel values are between 0 and 250 inclusive. The pixel value of 255 is NoData.
The Cloud Optimized GeoTIFF (COG) available for download:
Band 1: Dominant driver classification (categorical) over the time period
1 = Permanent agriculture
2 = Hard commodities
3 = Shifting cultivation
4 = Logging
5 = Wildfire
6 = Settlements and infrastructure
7 = Other natural disturbances
Band 2: Permanent agriculture - probability percentage (numeric) scaled over [0-250]
Band 3: Hard commodities - probability percentage (numeric) scaled over [0-250]
Band 4: Shifting cultivation - probability percentage (numeric) scaled over [0-250]
Band 5: Logging - probability percentage (numeric) scaled over [0-250]
Band 6: Wildfire - probability percentage (numeric) scaled over [0-250]
Band 7: Settlements and infrastructure - probability percentage (numeric) scaled over [0-250]
Band 8: Other natural disturbances - probability percentage (numeric) scaled over [0-250]
This product shows the dominant driver in each 1 km cell over the entire period. It does not show multiple drivers if they occur in the same cell at smaller scales, nor does it detail the sequence of drivers if multiple occurred at different times within the period.
This product does not distinguish between the loss of natural forest and planted trees (e.g., plantations, tree crops, or agroforestry systems). While tree cover loss associated with the permanent agriculture, hard commodities, and settlements & infrastructure classes represent a close approximation of deforestation, they do not always represent the conversion of natural forests to other land uses and in some cases may represent loss of planted trees. Similarly, replacement of natural forest with wood fiber plantations is not distinguished from routine harvesting within existing plantations established before 2000, as these are both included in the logging class.
These data are limited in scope to attributing drivers to tree cover loss as mapped by the Hansen et al. (2013) tree cover loss product, and therefore the detection of loss is subject to the accuracy of that product. A full description of limitations is included in the publication.
Global map of the dominant driver of tree cover loss at 0.01\u00b0 resolution (~1km) for the period 2001-2024. This is the latest update to this dataset.
", "has_chart_views": false, "id": "f3e94082-d49a-4b23-a42f-249bcd0531a2", "is_approved": true, "isopen": true, "language": "en", "learn_more": "https://www.globalforestwatch.org/blog/data-and-tools/new-drivers-data-forest-loss", "license_id": "cc-by", "license_title": "Creative Commons Attribution", "license_url": "http://www.opendefinition.org/licenses/cc-by", "maintainer": null, "maintainer_email": null, "maintainers": [{"name": "Michelle Sims", "email": "michelle.sims@wri.org"}, {"name": "Radost Stanimirova", "email": "radost.stanimirova@wri.org"}], "metadata_created": "2026-04-22T18:49:46.911265", "metadata_modified": "2026-04-29T19:16:20.354157", "methodology": "These data were produced by the World Resources Institute and Google DeepMind. The data were developed using a global neural network model (ResNet) trained on a set of 6,955 samples collected through visual interpretation of very high-resolution satellite imagery. The model used satellite imagery (Landsat 7 & 8, Sentinel-2) and ancillary data to classify the seven driver categories.
Overall accuracy of the model is 90.5%, with regional accuracies varying from 82.8% in Southeast Asia to 94.1% in Asia. Global per class producer\u2019s and user\u2019s accuracy are highest for the permanent agriculture, logging, and wildfire classes (over 90%), and generally lower for rarer classes, such as hard commodities, settlements and infrastructure, and other natural disturbances. A full description of the methods and accuracy statistics are available in the publication.
", "name": "dominant-drivers-of-tree-cover-loss-at-1km-v1-2", "notes": "This dataset contains the dominant driver of tree cover loss from 2001-2024. A driver is defined as the direct cause of tree cover loss, and can include both temporary disturbances (natural or anthropogenic) or permanent loss of tree cover due to a change to a non-forest land use (e.g., deforestation). The dominant driver is defined as the direct driver that caused the majority of tree cover loss within each 1 km cell over a time period.
Classes of drivers are defined as follows:
Permanent agriculture: Long-term, permanent tree cover loss for small- to large-scale agriculture.
Hard commodities: Loss due to the establishment or expansion of mining or energy infrastructure.
Shifting cultivation: Tree cover loss due to small- to medium-scale clearing for temporary cultivation that is later abandoned and followed by subsequent regrowth of secondary forest or vegetation.
Logging: Forest management and logging activities occurring within managed, natural or semi-natural forests and plantations, often with evidence of forest regrowth or planting in subsequent years.
Wildfire: Tree cover loss due to fire with no visible human conversion or agricultural activity afterward. Fires may be started by natural causes (e.g. lightning) or may be related to human activities (accidental or deliberate).
Settlements and infrastructure: Tree cover loss due to expansion and intensification of roads, settlements, urban areas, or built infrastructure (not associated with other classes).
Other natural disturbances: Tree cover loss due to other non-fire natural disturbances (e.g., landslides, insect outbreaks, river meandering). If loss due to natural causes is followed by salvage or sanitation logging, it is classified as logging.
The data is available for download as a global raster file with 8 bands, 8-bit pixel values, and 0.01-degree spatial resolution. The first band is a classification of the dominant driver for the full time period (2001-2024). The pixel values 1 through 7 correspond to the seven classifications of drivers and 255 corresponds to NoData. The other seven bands contain the probability of each driver. Probabilities should be interpreted as a relative confidence between the classes, rather than a true estimate of the likelihood of each class.
The pixel values for the probability bands should NOT be interpreted as an integer percentage (i.e. 1 means 1%, 2 means 2%, ... is false). Instead more dynamic range is provided and the percentage contribution between 0%-100% is quantized into 250 levels each corresponding to a 0.4% increment rather than a 1% increment. The formula that can be used to calculate the percentage probability for a driver is driver_probability_pct = uint8(pixel_value) * 0.4 . Valid pixel values are between 0 and 250 inclusive. The pixel value of 255 is NoData.
The Cloud Optimized GeoTIFF (COG) available for download:
Band 1: Dominant driver classification (categorical) over the time period
1 = Permanent agriculture
2 = Hard commodities
3 = Shifting cultivation
4 = Logging
5 = Wildfire
6 = Settlements and infrastructure
7 = Other natural disturbances
Band 2: Permanent agriculture - probability percentage (numeric) scaled over [0-250]
Band 3: Hard commodities - probability percentage (numeric) scaled over [0-250]
Band 4: Shifting cultivation - probability percentage (numeric) scaled over [0-250]
Band 5: Logging - probability percentage (numeric) scaled over [0-250]
Band 6: Wildfire - probability percentage (numeric) scaled over [0-250]
Band 7: Settlements and infrastructure - probability percentage (numeric) scaled over [0-250]
Band 8: Other natural disturbances - probability percentage (numeric) scaled over [0-250]
This modeled data builds on the Dataset of U.S. School Bus Fleets constructed by the Electric School Bus Initiative.
\nEstimating the climate impact of a school bus, diesel or electric, is performed by estimating GHG emissions associated with its operation in a straightforward process.\nFor such impacts the effects are not dependent on the location of the release of the GHGs.\nIn contrast the health impacts of pollutants such as fine particulate matter are highly location dependent.\nThe complexity and value of this dataset is primarily in the population health impact estimates and methodological approach.
\nThe data is the result of a reverse source influence model that estimates the impact of emissions associated with school buses in various counties in the contiguous United States.\nReverse source influence modeling uses an augmented version of a photochemical air quality model (AQM), known as an adjoint model, that traces the overall health impact on the entire population back to individual polluting sources at all locations.\nAs described in the methodology document the model starts from the overall population health impact as characterized by the combination of a regular AQM and an epidemiological model then attributes these impacts to emissions from each school bus at each location (i.e., each US county).\nInformation about input data and models to the impact model including meteorological and emission models is fully described in the WRI Technical Note.
", "name": "usa-school-bus-fleet-electrification-impacts", "notes": "This dataset is the output of a model that quantifies the health and climate impacts of replacing aging diesel school buses with either new diesel or new electric buses, incorporating both upstream and tailpipe emissions at the US county level.\nUsing reverse source influence modeling, the research estimates the societal benefits of various fleet renewal scenarios.\nKey questions include which regions would benefit most from electrification and what the national emissions would be under different replacement strategies.
\nThe findings are intended to inform policymakers, school districts, manufacturers, and utilities by providing monetized estimates of health and climate impacts, enabling data-driven decisions about school bus fleet modernization.
\nThe datasets produced include the following:
\nThe data is distributed as a set of Microsoft Excel files breaking the data down by fleet characteristics related to the year of the diesel school bus.\nThe county-level fleet is available for each year or in clusters of years corresponding to pre-2000, pre-2007, and post-2010.
", "num_resources": 1, "num_tags": 11, "open_in": "[{\"title\":\"Electric School Bus Initiative Dashboard\",\"url\":\"https://electricschoolbusinitiative.org/electric-school-bus-data-dashboard\"}]", "organization": {"id": "164c340b-5cf2-49da-98f9-024fdb7e0e42", "name": "electric-school-bus-initiative", "title": "Electric School Bus Initiative", "type": "organization", "description": "WRI's Electric School Bus Initiative accelerates the adoption of electric school buses to deliver on improved health and environmental outcomes in communities across the United States of America.", "image_url": "students-boarding-school-bus-m3277f.jpg", "created": "2024-09-24T18:14:31.852003", "is_organization": true, "approval_status": "approved", "state": "active"}, "owner_org": "164c340b-5cf2-49da-98f9-024fdb7e0e42", "private": false, "project": "", "rw_dataset": false, "rw_id": "", "short_description": "Quantification in dollar value of the impacts of transitioning school bus operations from diesel to electric for each county in the contiguous United States.", "spatial": "null", "spatial_address": "United States", "spatial_type": "address", "state": "active", "technical_notes": "https://doi.org/10.46830/writn.24.00024", "title": "Societal Health and Climate Benefits Associated With Transitioning the US School Bus Fleet From Diesel to Electric", "type": "dataset", "update_frequency": "not_planned", "version": null, "visibility_type": "public", "wri_data": true, "groups": [{"description": "Data concerning urban spaces and systems.", "display_name": "Cities", "id": "80539133-5aaa-4257-9aeb-fe6f56e2837f", "image_display_url": "https://datasets.wri.org/private-admin/uploads/group/children-biking-calle-colombia-2-qy6gqa.jpeg", "name": "cities", "title": "Cities", "type": "group"}], "resources": [{"cache_last_updated": null, "cache_url": null, "created": "2025-10-21T10:31:59.425043", "datastore_active": false, "description": "MS Excel documents (23 MB).", "format": "ZIP", "hash": "", "id": "f072f149-9953-4dce-b09e-57130a482e66", "last_modified": "2025-10-21T10:31:59.207920", "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-03-11T17:14:09.888586", "mimetype": "application/zip", "mimetype_inner": null, "name": "Zipfile", "new": false, "not_downloadable": false, "package_id": "de8149ed-9fba-42c1-999c-881fd580cb51", "position": 0, "resourceId": "f072f149-9953-4dce-b09e-57130a482e66", "resource_type": null, "size": 23936665, "state": "active", "title": "Zipfile", "type": "upload", "url": "https://datasets.wri.org/private-admin/dataset/de8149ed-9fba-42c1-999c-881fd580cb51/resource/f072f149-9953-4dce-b09e-57130a482e66/download/usa-school-bus-fleet-electrification-impacts.zip", "url_type": "upload"}], "tags": [{"display_name": "Air Quality", "id": "efe8706f-029d-4cfa-843b-4154d7d1bb3a", "name": "Air Quality", "state": "active", "vocabulary_id": null}, {"display_name": "Diesel", "id": "6cd10db1-3045-4f38-8a58-69044c73fffd", "name": "Diesel", "state": "active", "vocabulary_id": null}, {"display_name": "ESBs", "id": "01b47723-1faa-4ce1-864d-dae8827fd128", "name": "ESBs", "state": "active", "vocabulary_id": null}, {"display_name": "EV", "id": "93addc43-9fea-4a30-914b-f257899bc3ad", "name": "EV", "state": "active", "vocabulary_id": null}, {"display_name": "Electric School Buses", "id": "bf4f785e-6fc3-4e30-a36e-e039a2794039", "name": "Electric School Buses", "state": "active", "vocabulary_id": null}, {"display_name": "Electric Vehicles", "id": "a57e2462-36f7-4bb4-aece-bea5a5033e46", "name": "Electric Vehicles", "state": "active", "vocabulary_id": null}, {"display_name": "Emissions", "id": "57d47789-40c8-4bf1-976e-df257daa0a31", "name": "Emissions", "state": "active", "vocabulary_id": null}, {"display_name": "Exhaust", "id": "5ff8e5e4-10bd-4c94-a8b8-bae22cf25a82", "name": "Exhaust", "state": "active", "vocabulary_id": null}, {"display_name": "Health", "id": "115186ca-97d7-40e0-af67-8fe73096f136", "name": "Health", "state": "active", "vocabulary_id": null}, {"display_name": "Public Health", "id": "0adf6178-9725-4943-8efa-d8908b36be64", "name": "Public Health", "state": "active", "vocabulary_id": null}, {"display_name": "School Buses", "id": "c5a3d4f1-a7b3-4db1-b54b-43a1c55bccf4", "name": "School Buses", "state": "active", "vocabulary_id": null}], "relationships_as_subject": [], "relationships_as_object": []}, {"approval_status": "approved", "author": null, "author_email": null, "authors": [{"name": "Leandro Parente", "email": "leandro.parente@opengeohub.org"}, {"name": "Lindsey Sloat", "email": "lindsey.sloat@wri.org"}, {"name": "... and 18 others", "email": null}], "cautions": "These maps are produced through a complex data pipeline that uses satellite remote sensing products and labeled training data as critical inputs.\nOver 2.3-million reference samples have been collected using very high resolution imagery and combined with the bi-monthly Landsat ARD2 collection, long-term MODIS temperature and water vapor data, geometric temperature data, global terrain and elevation data, and maps of distance to key land structures such as roads and waterways.\nThe reference labels and Earth observation data are used for the training of spatiotemporal machine learning models with one model per land cover class.\nThese models are applied in a prediction step to create probabalistic estimates of land cover which are then balanced and harmonized to produce maps of the dominant grassland class.
\nThe full methodology is detailed in the manuscript and should be consulted to best understand specific definitions used in this data product.
\nThe software supporting the machine learning model development and inference is available on GitHub: https://github.com/wri/global-pasture-watch.
", "name": "grassland-dynamics", "notes": "This product maps global grassland dynamics annually for 2000-2022 at 30 m spatial resolution.\nThe dataset showing the spatiotemporal distribution of cultivated and natural/semi-natural grassland classes was produced by using bi-monthly aggregates of GLAD Landsat ARD-2 image archive, accompanied by climatic, landform and proximity covariates, spatiotemporal machine learning (per-class random forest) and over 2.3-million reference samples visually interpreted in very high resolution imagery.\nThe suggested uses of data include (1) integration with other compatible land cover products and (2) tracking the intensity and drivers of conversion of land to cultivated grasslands and from natural / semi-natural grasslands into other land use systems.
\nThe mapped grassland extent includes any land cover type which contains at least 30% of dry or wet low vegetation dominated by grasses and forbs (less than 3 meters) and a:
\nThe grassland extent is classified into two classes:
\nThe dataset is organized in 69 global mosaics (23 years for each time series) in COG (Cloud Optimized GeoTIFF) format, WGS84 Coordinate Systems (EPSG:4326) and pixel size equal to 0.00025 degrees, including:
\nThese COGs are available in the OpenLandMap STAC: https://stac.openlandmap.org/gpw_ggc-30m/collection.json
\nThe data is available for visualization and analysis within Google Earth Engine by using the following ImageCollections:
\nA Google Earth Engine Application is also available for inspecting the data and investigating custom probability thresholds for class selection: https://global-pasture-watch.projects.earthengine.app/view/ggc-30m.
\nThe grassland dynamics maps are created by the Global Pasture Watch (GPW) research consortium initiated by the Land & Carbon Lab.\nGPW consists of experts from the World Resources Institute (WRI), OpenGeoHub Foundation, the Image Processing and GIS Laboratory at the Federal University of Goi\u00e1s (LAPIG/UFG), the International Institute for Applied Systems Analysis (IIASA), the German Center for Integrative Biodiversity Research (iDiv), Cornell University; and the Global Land Analysis and Discovery laboratory of the University of Maryland (GLAD).
", "num_resources": 5, "num_tags": 3, "open_in": "[{\"title\":\"Google Earth Engine\",\"url\":\"https://developers.google.com/earth-engine/datasets/publisher/global-pasture-watch\"}]", "organization": {"id": "ad461dbd-f646-4080-94a9-c10005c42a43", "name": "land-carbon-lab", "title": "Land & Carbon Lab", "type": "organization", "description": "WRI's Land & Carbon Lab creates datasets to support critical monitoring and decision making needs surrounding land use planning, carbon accounting, and other technical facets of human-land interactions.", "image_url": "1724353785-cover-ethiopia-mosaic-oqiid6.avif", "created": "2024-09-25T11:07:47.708772", "is_organization": true, "approval_status": "approved", "state": "active"}, "owner_org": "ad461dbd-f646-4080-94a9-c10005c42a43", "private": false, "project": "", "rw_dataset": false, "rw_id": "", "short_description": "Global maps of grassland class created from satellite imagery with 30-meter spatial resolution and annual temporal resolution.", "spatial": "null", "spatial_address": "Global", "spatial_type": "global", "state": "active", "technical_notes": "https://doi.org/10.1038/s41597-024-04139-6", "title": "Annual 30-m maps of global grassland class and extent (2000\u20132022)", "type": "dataset", "update_frequency": "not_planned", "version": null, "visibility_type": "public", "wri_data": true, "groups": [{"description": "Data concerning land use, land cover, and terrestrial ecosystem dynamics along with the human and environmental drivers of our food, forests, and water systems.", "display_name": "Land", "id": "c1427a08-9de5-4182-9c05-b6a61fda3127", "image_display_url": "https://datasets.wri.org/private-admin/uploads/group/land-peru-amazon-river-jrjo7f.jpg", "name": "land", "title": "Land", "type": "group"}], "resources": [{"cache_last_updated": null, "cache_url": null, "created": "2024-12-16T18:21:53.213844", "datastore_active": false, "description": "Global raster with 3 classes - other land cover (0), cultivated (1), or natural/semi-natural grassland (2) based on balanced and harmonized thresholds for the grassland class probabilities.", "format": "Layer", "hash": "", "id": "16af6998-0343-4ea0-b2f0-a7d6424ba823", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-02-10T16:23:48.474549", "mimetype": null, "mimetype_inner": null, "name": "Cultivated and natural/semi-natural grassland map", "new": false, "not_downloadable": false, "package_id": "ab526ddc-3954-438a-9a04-2fbb057fa53c", "position": 0, "resourceId": "16af6998-0343-4ea0-b2f0-a7d6424ba823", "resource_type": null, "rw_id": "8656f38d-f826-4660-b83f-f5c3eedbd430", "size": null, "state": "active", "title": "Cultivated and natural/semi-natural grassland map", "type": "reference-layer", "url": "https://api.resourcewatch.org/v1/dataset/f127ef4d-91ae-4398-b4d7-57a157b31990/layer/8656f38d-f826-4660-b83f-f5c3eedbd430", "url_type": "reference-layer"}, {"cache_last_updated": null, "cache_url": null, "created": "2024-12-16T18:21:53.213852", "datastore_active": false, "description": "Probability between 0% and 100% that the land is natural/semi-natural grassland or pasture.", "format": "Layer", "hash": "", "id": "32417676-ff67-4113-8cf2-67d142227914", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-02-10T16:23:48.474666", "mimetype": null, "mimetype_inner": null, "name": "Natural/semi-natural probability", "new": false, "not_downloadable": false, "package_id": "ab526ddc-3954-438a-9a04-2fbb057fa53c", "position": 1, "resourceId": "32417676-ff67-4113-8cf2-67d142227914", "resource_type": null, "rw_id": "31f4b063-0a43-4cde-a178-8470b514fb51", "size": null, "state": "active", "title": "Natural/semi-natural probability", "type": "reference-layer", "url": "https://api.resourcewatch.org/v1/dataset/52ba9748-107f-4b2a-94ec-e9d5f16598ba/layer/31f4b063-0a43-4cde-a178-8470b514fb51", "url_type": "reference-layer"}, {"cache_last_updated": null, "cache_url": null, "created": "2024-12-16T18:21:53.213855", "datastore_active": false, "description": "Probability between 0% and 100% that the land is cultivated grassland or pasture.", "format": "Layer", "hash": "", "id": "64fe432d-4546-4808-b330-6c97287d6d98", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-02-10T16:23:48.474741", "mimetype": null, "mimetype_inner": null, "name": "Cultivated probability", "new": false, "not_downloadable": false, "package_id": "ab526ddc-3954-438a-9a04-2fbb057fa53c", "position": 2, "resourceId": "64fe432d-4546-4808-b330-6c97287d6d98", "resource_type": null, "rw_id": "c3b7723f-e1e4-4ae2-b7af-86b0115d0dd1", "size": null, "state": "active", "title": "Cultivated probability", "type": "reference-layer", "url": "https://api.resourcewatch.org/v1/dataset/4849f6a8-f43a-432a-96a0-aebfef74aeea/layer/c3b7723f-e1e4-4ae2-b7af-86b0115d0dd1", "url_type": "reference-layer"}, {"cache_last_updated": null, "cache_url": null, "created": "2024-12-16T18:21:53.213858", "datastore_active": false, "description": "This data product has been archived with Zenodo and you are able to access the GeoTIFFs using this link.", "format": "", "hash": "", "id": "cac29106-e4de-4815-8ef7-8dcd421aad2e", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-02-10T16:23:48.474822", "mimetype": null, "mimetype_inner": null, "name": "Cloud Optimized GeoTIFFs (COGs)", "new": false, "not_downloadable": false, "package_id": "ab526ddc-3954-438a-9a04-2fbb057fa53c", "position": 3, "resourceId": "cac29106-e4de-4815-8ef7-8dcd421aad2e", "resource_type": null, "size": null, "spatial_address": null, "spatial_coordinates": null, "spatial_geom": null, "spatial_type": null, "state": "active", "title": "Cloud Optimized GeoTIFFs (COGs)", "type": "link", "url": "https://doi.org/10.5281/zenodo.13890401", "url_type": "link"}, {"asset_id": "ee.ImageCollection(\"projects/global-pasture-watch/assets/ggc-30m/v1/grassland_c\")", "asset_type": "raster", "cache_last_updated": null, "cache_url": null, "created": "2026-02-10T16:21:13.165000", "datastore_active": false, "description": "Global raster with 3 classes - other land cover (0), cultivated (1), or natural/semi-natural grassland (2) based on balanced and harmonized thresholds for the grassland class probabilities.", "format": "", "hash": "", "id": "e0b7da86-4945-4ba3-84ad-ff0757e93af7", "last_modified": null, "layerObj": null, "layerObjRaw": null, "metadata_modified": "2026-02-10T16:23:48.477610", "mimetype": null, "mimetype_inner": null, "new": false, "not_downloadable": false, "package_id": "ab526ddc-3954-438a-9a04-2fbb057fa53c", "position": 4, "resourceId": "e0b7da86-4945-4ba3-84ad-ff0757e93af7", "resource_type": null, "schema": {"value": []}, "size": null, "state": "active", "title": "GPW Annual Dominant Class", "type": "gee-asset", "url": "", "url_type": "gee-asset"}], "tags": [{"display_name": "Global Pasture Watch Consortium", "id": "230efe08-344a-463e-962c-76523382326a", "name": "Global Pasture Watch Consortium", "state": "active", "vocabulary_id": null}, {"display_name": "Raster", "id": "02c5322e-f55b-4f58-b473-3f7cf2356a8d", "name": "Raster", "state": "active", "vocabulary_id": null}, {"display_name": "STAC", "id": "0ada3f2d-f22a-411c-89cd-6474ebf57cf5", "name": "STAC", "state": "active", "vocabulary_id": null}], "relationships_as_subject": [], "relationships_as_object": []}], "sort": "score desc, metadata_modified desc", "search_facets": {}}}