Sep 26, 2025
Designing territories across a city, region, or country often means you need a stable, comparable baseline year to year.
Overview
LandScan is a widely used global gridded population dataset developed by Oak Ridge National Laboratory (ORNL). It models an ambient population—an estimate of where people are present throughout the day (daytime + nighttime)—disaggregated from census or administrative-level totals into fine spatial grids. LandScan is often used in risk, planning, emergency response, and infrastructure analyses because it offers consistent annual updates and spatial detail beyond coarse census units.
Methodology & Spatial Characteristics
Base resolution: ~30 arc-seconds (~1 km) globally.
Dasymetric weighting: Allocates census totals to cells using roads, land cover, slope, night-time lights, and other covariates.
Normalization: Grid sums match census or projection totals.
Ambient model: Reflects daily presence, not just residence.
LandScan HD: In select regions, ~3 arc-seconds (~90 m) grids for day/night distributions.
Strengths & Use Cases
Consistency: Annual, globally consistent updates support year-to-year comparability.
Ambient insight: Useful for risk, commuting, service, and infrastructure analyses.
Global coverage: Fills data gaps where censuses are sparse.
Limitations & Caveats
Resolution ceiling: Smoothing at ~1 km; neighborhood-level variation lost.
Model assumptions: Proxy weights may misallocate population in atypical settings.
Temporal lag: Census inputs may be outdated.
Ambient mismatch: Not directly comparable to residential-only counts.
Related Resources
Need More Help?
If you run into issues, please contact us.
