High-resolution population estimation using household survey data and building footprints

TL;DR

This paper presents a Bayesian hierarchical model that combines household survey data and building footprints to produce high-resolution, up-to-date population estimates in regions with outdated census data, demonstrated in the Democratic Republic of the Congo.

Contribution

The study introduces a novel Bayesian model integrating survey data and spatial building information for detailed population estimation in data-scarce settings.

Findings

Model achieved R^2 of 0.79 for population totals

Perfect fit (R^2=1.00) for age and sex proportions

Effective in regions with outdated census data

Abstract

The national census is an essential data source to support decision-making in many areas of public interest. However, this data may become outdated during the intercensal period, which can stretch up to several decades. We developed a Bayesian hierarchical model leveraging recent household surveys with probabilistic sampling designs and building footprints to produce up-to-date population estimates. We estimated population totals and age and sex breakdowns with associated uncertainty measures within grid cells of approximately 100m in five provinces of the Democratic Republic of the Congo, a country where the last census was completed in 1984. The model exhibited a very good fit, with an R^2 value of 0.79 for out-of-sample predictions of population totals at the microcensus-cluster level and 1.00 for age and sex proportions at the province level. The results confirm the benefits of combining household surveys and building footprints for high-resolution population estimation in countries with outdated censuses.