Back to Programme

An Efficient Geospatial Tool for Creating and Updating Census Enumeration Areas and National Sampling Frames

Sarchil Qader (WorldPop/University of Southampton) - United Kingdom
Andrew Harfoot (GeoData/University of Southampton) - United Kingdom
Andrew Tatem (WorldPop/University of Southampton) - United Kingdom

Keywords: preEA, Geospatial, Enumeration Areas, Census, National Sampling Frame


Abstract

Accurate and reliable Enumeration Area (EA) demarcation and national sampling frames are critical for the success of censuses and national health and socio-economic surveys. However, in many low-income or conflict-affected countries, EA demarcations are often incomplete, outdated, or nonexistent. Even in stable, middle- and high-income countries, the process of creating and updating EAs remains a complex challenge. Traditionally, manual methods are employed to delineate and update national sampling frames, a resource-intensive process that can take several years.
In response to these challenges, WorldPop, in collaboration with multiple National Statistical Offices (NSOs) and international organizations, has developed an automated tool (preEA) designed to efficiently generate and update national sampling frames. This tool utilizes high-resolution gridded population and settlement data, GPS household locations, building footprints, and publicly available natural, man-made, and administrative boundaries. The automation process employs a ‘split and merge’ methodology, drawing on concepts from image processing, specifically image segmentation using mathematical morphology (e.g., watershed and waterfall algorithms).
Initial applications of the tool were conducted in Nigeria, Benin, Colombia, Armenia, the Democratic Republic of Congo (DRC), Burkina Faso, Somalia, Paraguay, Cameroon, Togo, Zimbabwe, Niger, and Guinea. Results demonstrate that the generated EA boundaries align with international standards, incorporating visible ground features such as roads and waterways, adhering to administrative boundaries, and avoiding geometric errors. Field tests in DRC, Colombia, and Benin further assessed the tool's utility in preparing for national population and housing and economic censuses. The outcomes from these tests confirmed that the tool met the expectations of the NSOs in all three countries, leading to the adoption of the tool for census cartography preparation.
In addition to its role in EA demarcation, the preEA tool package offers critical metrics (e.g., hard-to-account indicators) necessary for operational planning, ensuring adequate staffing and equipment for on-the-ground activities.