Gridded population counts from WorldPop's Individual Countries 2020 UN-adjusted datasets (constrained and unconstrained variants), 234 countries/territories, 3 arc-second (~100 m), as Cloud-Optimized GeoTIFFs
WorldPop 2020 gridded population counts for individual countries, adjusted to UN national estimates, in constrained and unconstrained variants at ~100 m. The units are number of people per pixel with country totals adjusted to match the corresponding official United Nations population estimates (2019 Revision of World Population Prospects). The mapping approach is Random Forest-based dasymetric redistribution. The unconstrained variant distributes population over all land grid squares; the constrained variant restricts it to areas mapped as containing built settlements.
See also the source dataset page and the WorldPop constrained vs unconstrained methods note for methodology and caveats.
Storage for this dataset is generously provided by Source Cooperative, a Radiant Earth project.
collection.geoparquet — stac-geoparquet export of the collection's Items: one row per item, geometry plus asset hrefs<iso3>/ — one directory per country, keyed by lowercase ISO 3166-1 alpha-3 code*_constrained.tif / *_unconstrained.tif — the two population count variantsBrowse or download with the AWS CLI:
Each item's odc:region_code holds its country code. To subset by area instead, pass
bbox=[minx, miny, maxx, maxy] to both search_sync and load.
CC-BY-4.0. Attribute WorldPop as producer when redistributing or publishing derived work, citing the DOIs above:
This dataset was prepared and published through the WFP Humanitarian Data Cube (HDC).
It is also served live from the HDC STAC API:
wp_pop_icunadj.
For questions, corrections, or data requests, contact hq.geospatial@wfp.org.