Cognised existence: Population is the measured quantity and density of inhabitants across space and time.
Question: What is the population count, density, and demographic profile in this area?
Population is typically represented as raster grids or statistical polygons, not as settlement place points.
Realisations
Global population grids
Global gridded datasets estimate population counts and density on regular cells (for example 1 km or 100 m depending on product/version).
Representative products:
| Product | Geometry | Typical Resolution | Scope |
|---|---|---|---|
| GHSL Population Grid (GHS-POP) | Raster grid | 250 m to 1 km | Global |
| WorldPop | Raster grid | ~100 m (country dependent) | Global (modelled) |
| GPWv4 (CIESIN) | Raster grid | ~1 km | Global |
EU population grids
For Europe, population is commonly distributed in harmonized grid systems for comparable regional analysis.
Representative products:
| Product | Geometry | Typical Resolution | Scope |
|---|---|---|---|
| GEOSTAT population grid | Grid polygon / raster equivalent | 1 km | Europe |
| National statistical population grids | Grid polygon / raster | 100 m to 1 km | Country-specific |
Census and administrative statistics
Population can also be represented as areal totals linked to administrative units (municipality, region, census tract, etc.).
Python load (example pattern)
import geopandas as gpd
import rasterio
# Raster grid example (population count per cell)
with rasterio.open("population_grid.tif") as src:
pop_grid = src.read(1)
transform = src.transform
# Polygon example (population per statistical unit)
admin_pop = gpd.read_file("population_by_admin_units.gpkg")
admin_pop["pop_density"] = admin_pop["population"] / admin_pop.to_crs(3857).area * 1_000_000Geometry Representations
| Rep ID | Source Dataset | Geometry Type | Native CRS | Suitable For | Not Suitable For |
|---|---|---|---|---|---|
population_grid_global | GHSL / WorldPop / GPW | Raster grid | Product-specific | Density surfaces, exposure modelling, regional comparisons | Household-level counts, legal reporting |
population_grid_eu | GEOSTAT / national EU grids | Grid polygon or raster | Product-specific | Harmonized European population comparisons, density analysis | Fine-grained individual-level inference |
population_admin_units | Census by admin area | Polygon | National statistical CRS | Official reporting, demographic indicators by area | Within-unit spatial heterogeneity |
Limitations
- Gridded population products are modelled estimates, not exact headcounts per cell.
- Census polygons introduce the modifiable areal unit problem (MAUP).
- Temporal mismatch across products is common; always check reference year.
- Settlement point datasets should be treated as populated-area proxies, not full population measures.
Related Leaves
Realised By Links
- INSPIRE Population Distribution (classification)
- UN-GGIM Population Distribution (classification)