1,720,992 research outputs found
Binary classification of the Kinshasa and Bandundu provinces in the Democratic Republic of the Congo — settled versus non settled
No full textThis dataset was created based on a settlement layer produced by the Oak Ridge National Laboratory using feature extraction from high-resolution imagery for population modelling work undertaken in the Kinshasa and Kongo-Central provinces in the Democratic Republic of the Congo. The settlement layer consists of settlement polygons of approximately 7 meters resolution. The polygons were rasterized based on a reference grid with a resolution of 3 arc-seconds, approximately 90 meters. The presence of at least one settlement polygon designated a settled cell. We thank the Oak Ridge National Laboratory and the Bill and Melinda Gates Foundation for the support.
We would also like to extend our gratitude to Eric M. Webber and Amy N. Rose at the Oak Ridge National Laboratory and Io Blair-Freese at the Bill and Melinda Gates Foundation.</span
A bottom-up population modelling approach to complement the population and housing census
No full textPopulation and housing censuses provide essential demographic information for local, national and international decision-making and response. However, census data in the most vulnerable countries are often outdated or partial because political instability, conflict and natural disasters prevent a nationwide enumeration. The bottom-up modelling approach complements outdated or incomplete census data by estimating population counts and age/sex structures in grid cells of about 100 m using required population data on a set of fully enumerated locations and auxiliary geospatial covariates. We present the modelling effort in the Democratic Republic of Congo - the last census was conducted in 1984 - and in Burkina Faso - the last census was conducted in 2020 but covered only 70% of the country. Both models showed good predictive performance, denoted by R2 values of 0.73 and 0.63 for the respective out-of-sample predictions of population counts. The resulting bottom-up and gridded population estimates are currently used for census support and humanitarian response in both countries. This work has highlighted the flexibility of the bottom-up modelling approach, in terms of input population data, model specification and aggregation of population estimates to support specific use cases
A grid-based sample design framework for household surveys
No full textTraditional sample designs for household surveys are contingent upon the availability of a representative primary sampling frame. This is defined using enumeration units and population counts retrieved from decennial national censuses that can become rapidly inaccurate in highly dynamic demographic settings. To tackle the need for representative sampling frames, we propose an original grid-based sample design framework introducing essential concepts of spatial sampling in household surveys. In this framework, the sampling frame is defined based on gridded population estimates and formalized as a bi-dimensional random field, characterized by spatial trends, spatial autocorrelation, and stratification. The sampling design reflects the characteristics of the random field by combining contextual stratification and proportional to population size sampling. A nonparametric estimator is applied to evaluate the sampling design and inform sample size estimation. We demonstrate an application of the proposed framework through a case study developed in two provinces located in the western part of the Democratic Republic of the Congo. We define a sampling frame consisting of settled cells with associated population estimates. We then perform a contextual stratification by applying a principal component analysis (PCA) and k-means clustering to a set of gridded geospatial covariates, and sample settled cells proportionally to population size. Lastly, we evaluate the sampling design by contrasting the empirical cumulative distribution function for the entire population of interest and its weighted counterpart across different sample sizes and identify an adequate sample size using the Kolmogorov-Smirnov distance between the two functions. The results of the case study underscore the strengths and limitations of the proposed grid-based sample design framework and foster further research into the application of spatial sampling concepts in household surveys.</p
Input data and code supporting the cod_v2 population estimates
No full textThe model.zip file contains input data and code supporting the cod_v2 population estimates. The file modelData.RData provides the input data to the JAGS model and the file modelCode.R contains the source code for the model in the JAGS language. The files can be used to run the model for further assessments and as a starting point for further model development.
The data and the model were developed using the statistical software R version 4.0.2 (https://cran.r-project.org/bin/windows/base/old/4.0.2) and JAGS 4.3.0 (https://mcmc-jags.sourceforge.io), a program for analysis of Bayesian graphical models using Gibbs sampling, through the R package runjags 2.2.0 (https://cran.r-project.org/web/packages/runjags).</span
A grid-based sampling design for household surveys in the absence of actionable sampling frames
No full textHousehold surveys are a cost-effective data source for estimating health and demographic characteristics in low- and middle-income countries. These surveys involve sampling from a frame listing all the members of the population of interest. However, the listing needs to be complete, accurate, and up-to-date to draw samples that can be generalized to the entire population. A sampling frame generally consists of the enumeration areas established during the last national census, together with the associated population counts. However, these counts are often inaccurate because the national census is carried out on a decade basis. Outdated sampling frames are particularly critical in Africa, where five countries had their last census more than fifteen years ago. To tackle the absence of actionable sampling frames, we propose an original grid-based sampling design, embedding spatial sampling concepts into household surveys. We demonstrate our framework with a case study developed in the western part of the Democratic Republic of Congo (DRC). This country had its last national census in 1984, and the resulting sampling frame is still currently used in household surveys
High-resolution population mapping and estimation in the western part of the Democratic Republic of Congo
No full textAccording to the United Nations, the Democratic Republic of Congo (DRC) is the fourth most populous country in Africa. Yet, accurate demographic figures are currently unavailable because the last national census was conducted over three decades ago. This situation has shown to be critical to policymaking and intervention planning in the domain of public health. In the western part of the country, for instance, human African trypanosomiasis is highly endemic but accurate sub-national population figures are required for effective disease mapping. This work aims at producing gridded population estimates with a 100 meters spatial resolution for the five most western provinces of DRC. In doing so, we adopt a bottom-up modeling approach incorporating demographic data from recent micro-census surveys and various geospatial datasets in a Bayesian hierarchical framework. The resulting high-resolution population estimates and associated uncertainty measures are meant to inform spatial epidemiological applications in the region and foster the use of similar modeling approaches in other parts of the country
Modelled gridded population estimates Kasaï-Oriental Province in the Democratic Republic of Congo version 4.3
No full textThis dataset consists of gridded population estimates for the Kasaï-Oriental province in the Democratic Republic of Congo (DRC). It includes gridded population counts with model uncertainty measures and breakdowns in 40 age-sex groups at a spatial resolution of 3 arc-seconds, approximately 100-metre grid cells. The estimates are produced in a Bayesian hierarchical modelling framework by combining population and building count data collected in a microcensus survey, gridded building counts from settlement extent data, and other gridded geospatial covariates. The model estimates and predicts population and building counts based on the microcensus survey data using gridded settlement extent data as an essential model covariate. Although this method incorporates uncertainty in the population and building input data, other unaccounted sources of uncertainty are most likely present. These model-based population estimates can be considered as most accurately representing the year 2024. This time period corresponds to inputs used for Kasaï-Oriental.
The data were produced by the WorldPop Research Group at the University of Southampton as part of the GRID3 – Phase 2 Scaling project, with funding from the Gates Foundation (INV-044979). Project partners included GRID3 Inc, the Center for Integrated Earth System Information (CIESIN) within the Columbia Climate School at Columbia University, and WorldPop at the University of Southampton. The statistical model was designed, developed, and implemented by Gianluca Boo. Data processing was done by Ortis Yankey and Tom Abbott with additional support from Amy Bonnie and Heather Chamberlain. Methodological and project oversight was provided by Chris Nnanatu, Attila Lazar, and Andy Tatem. The microcensus survey data was collected during the GRID3 Mapping for Health project funded by Gavi, the Vaccine Alliance (RM 86720420A2). CIESIN prepared and shared the settlement extent data. The data has been clipped to GRID3-CIESIN health area extent (version 6.0) (CIESIN, 2025). The whole WorldPop group is acknowledged for overall support, particularly Chris Nnanatu, Attila Lazar, and Ortis Yankey for reviewing and providing thoughtful suggestions.
</span
Exploring uncertainty in canine cancer data sources through dasymetric refinement
Get PDFIn spite of the potentially groundbreaking environmental sentinel applications, studies of canine cancer data sources are often limited due to undercounting of cancer cases. This source of uncertainty might be further amplified through the process of spatial data aggregation, manifested as part of the modifiable areal unit problem (MAUP). In this study, we explore potential explanatory factors for canine cancer incidence retrieved from the Swiss Canine Cancer Registry (SCCR) in a regression modeling framework. In doing so, we also evaluate differences in statistical performance and associations resulting from a dasymetric refinement of municipal units to their portion of residential land. Our findings document severe underascertainment of cancer cases in the SCCR, which we linked to specific demographic characteristics and reduced use of veterinary care. These explanatory factors result in improved statistical performance when computed using dasymetrically refined units. This suggests that dasymetric mapping should be further tested in geographic correlation studies of canine cancer incidence and in future comparative studies involving human cancers.</p
Gridded maps of building patterns throughout sub-Saharan Africa, version 1.1
No full textGridded maps of building patterns throughout sub-Saharan Africa, version 1.1. Source of building footprints "Ecopia Vector Maps Powered by Maxar Satellite Imagery" © 2020.</span
The population seen from space: when satellite images come to the rescue of the census
Get PDFThe size of the population, the denominator of many statistical indicators, is crucial for public policy. National statistical offices organize the collection of this information, most often through a census. But what happens when parts of a country are not accessible to census enumerators? Today, spatial data extracted from satellite imagery offer high-resolution geographical information with complete coverage. When combined with a partial population count, they offer an unprecedented opportunity to estimate the size of the population in inaccessible areas. The spatial precision of these data also makes possible the production of a high-resolution gridded population estimate, an innovative data format at the intersection of geography and demography. Based on the case of Burkina Faso, this article analyses how, by dividing a country into 100 m by 100 m cells, a Bayesian hierarchical model can be used to estimate the population of areas with security challenges which could not be enumerated during the 2019 census. This gridding allows the resulting counts to be disaggregated using a statistical learning model, yielding unparalleled spatial precision in population estimates
- …
