Global Pasture Watch - Livestock reference samples based on multi-source sub-national census data (2000—2022)

Harmonized collection from various national agencies, FAOSTAT and existing livestock data sources. Data were downloaded, curated and standardized using an ontology framework, and later integrated with existing livestock polygonal data. For each polygon administrative unit name, reference year, livestock animal type and headcount were aassigned.

The dataset includes information for cattle, horses, goats and sheep is composed by:

• GPW: Time series of livestock census data obtained from various national agencies organized by Global Pasture Watch project.
• GLW: Subset of the input census polygons used for producing Gridded Livestock of the World v4 (Wisser & Cinardi, 2024)
• Malek et al. (2024): Including European census data of cattle and sheep headcounts for 2020 and 2010
• FAOSTAT: Used to backfill data gaps in small countries (total national area smaller than 226 thousands km2)

Processing pipeline implementes using arealDB, ontologics, tabshiftr, Pandas and GeoPandas are publicly available in https://github.com/wri/global-pasture-watch.

Related resources

• Maps of livestock headcount for 2000—2022 (FAOSTAT-adjusted):
  All animals
• Maps of livestock density for 2000—2022 (including prediction interval):
  Cattle  Goat   Sheep  Horse
• Livestock reference census data (2000–2022):
  CSV and GeoPackage files
• Maps potential land for livestock production for 2000—2022 (including production systems)
  Raster files
• Data catalogues:
  OpenLandMap STAC Google Earth Engine