A Bayesian approach to produce 100 m gridded population estimates using census microdata and recent building footprints.
A Bayesian approach to produce 100 m gridded population estimates using census microdata and recent building footprints.
This report describes a novel Bayesian statistical method that combines recent building footprints from Ecopia.AI and Maxar Technologies with publicly-available census microdata from IPUMS International to produce 100 m gridded population estimates for Ghana. The model was used to estimate total populations, populations within specific age-sex groups, number of households, people per household, and households per building. Bayesian estimates of uncertainty are provided with all parameter estimates. Supplementary files are included with input data and statistical model code in the Stan programming language. This method is generalizable to additional countries where IPUMS data and building footprints are available.
Human population, Bayesian statistics, bensus microdata, building footprints
University of Southampton
Leasure, Douglas
c025de11-3c61-45b0-9b19-68d1d37959cd
Tatem, Andrew
6c6de104-a5f9-46e0-bb93-a1a7c980513e
Bondarenko, Maksym
1cbea387-2a42-4061-9713-bbfdf4d11226
Leasure, Douglas
c025de11-3c61-45b0-9b19-68d1d37959cd
Tatem, Andrew
6c6de104-a5f9-46e0-bb93-a1a7c980513e
Bondarenko, Maksym
1cbea387-2a42-4061-9713-bbfdf4d11226
Leasure, Douglas and Tatem, Andrew
(2020)
A Bayesian approach to produce 100 m gridded population estimates using census microdata and recent building footprints.
University of Southampton
doi:10.5258/SOTON/WP00686
[Dataset]
Abstract
This report describes a novel Bayesian statistical method that combines recent building footprints from Ecopia.AI and Maxar Technologies with publicly-available census microdata from IPUMS International to produce 100 m gridded population estimates for Ghana. The model was used to estimate total populations, populations within specific age-sex groups, number of households, people per household, and households per building. Bayesian estimates of uncertainty are provided with all parameter estimates. Supplementary files are included with input data and statistical model code in the Stan programming language. This method is generalizable to additional countries where IPUMS data and building footprints are available.
This record has no associated files available for download.
More information
Published date: 17 November 2020
Keywords:
Human population, Bayesian statistics, bensus microdata, building footprints
Identifiers
Local EPrints ID: 445047
URI: http://eprints.soton.ac.uk/id/eprint/445047
PURE UUID: b9a25552-e05a-4ec0-ba41-b2fd5fd98c37
Catalogue record
Date deposited: 18 Nov 2020 17:31
Last modified: 06 May 2023 01:58
Export record
Altmetrics
Contributors
Creator:
Douglas Leasure
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics