GPWv4: Global Population Density Grids

2000 to 2020 population density estimates are provided as global raster data. The data is derived from national censuses and registers, using a proportional allocation algorithm across approximately 13.5 million administrative units. Rasters are available at multiple resolutions, from 30 arc-seconds (~1 km) up to 1 degree, to support different processing needs.

Use Cases

• Modeling human exposure to environmental hazards based on gridded population density.
• Integrating demographic data with other geospatial layers for sustainable development planning.
• Analyzing population distribution trends from 2000 to 2020 using the multi-temporal rasters.
• Conducting global or regional analyses using aggregated data at coarser resolutions for faster processing.

Strengths

• Provides a consistent time series of five data points (2000, 2005, 2010, 2015, 2020).
• Utilizes a detailed input of approximately 13.5 million administrative units for gridding.
• Offers data at multiple spatial resolutions (30 arc-second to 1 degree) for flexibility.

Limitations

• Metadata completeness is low; column names, row counts, and license information are not provided.
• There is a conflict in the 'last updated' date between sources (2026-06-16 on Dataverse vs. 2020-07-01 on NASA Earthdata).

Provenance

Source

Center for International Earth Science Information Network - CIESIN - Columbia University

Collection Method

A proportional allocation gridding algorithm applied to census and population register counts.

Time Range

2000-2020

Freshness

2026-06-16

Geography

Global