A data fusion approach to the estimation of temporary populations: An application to Australia

• Published in: PloS one Vol. 16; no. 11; p. e0259377
• Main Authors: Charles-Edwards, Elin, Corcoran, Jonathan, Loginova, Julia, Panczak, Radoslaw, White, Gentry, Whitehead, Alexander
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 11.11.2021; Public Library of Science (PLoS)
• Subjects: Australia; Computer and Information Sciences; Data integration; Data mining; Data sources; Datasets; Earth Sciences; Emergency preparedness; Engineering and Technology; Estimates; Estimation; Ethics; Evaluation; Humans; Information management; Medicine and Health Sciences; Modelling; Models, Statistical; Monthly variations; People and Places; Population (statistical); Population biology; Population Density; Population statistics; Privacy; Research and Analysis Methods; Seasonal variations; Social networks; Social Sciences; Spatio-Temporal Analysis; Statistics; Time Factors; Australia; Queensland Australia; St Lucia
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0259377

This study establishes a new method for estimating the monthly Average Population Present (APP) in Australian regions. Conventional population statistics, which enumerate people where they usually live, ignore the significant spatial mobility driving short term shifts in population numbers. Estimates of the temporary or ambient population of a region have several important applications including the provision of goods and services, emergency preparedness and serve as more appropriate denominators for a range of social statistics. This paper develops a flexible modelling framework to generate APP estimates from an integrated suite of conventional and novel data sources. The resultant APP estimates reveal the considerable seasonality in small area populations across Australia's regions alongside the contribution of domestic and international visitors as well as absent residents to the observed monthly variations. The modelling framework developed in the paper is conceived in a manner such that it can be adapted and re-deployed both for use with alternative data sources as well as other situational contexts for the estimation of temporary populations.