Do Remote Sensing Population Datasets Provide a Realistic Picture of the Spatial Distribution of Populations in the Arctic? Empirical Comparison with Data from the Swedish Population Register.

Thibaud Bureau du Colombier, Université de Strasbourg

Population datasets built from remote sensing data are a great tool for examining population distribution based on several covariates such as land use, night lighting or water bodies. Top-down datasets aim to provide a realistic spatial distribution of population change over time in many countries, to match official data. These promising datasets could enable researchers to better understand the past evolution of population distribution and its mechanisms in different contexts. This knowledge would enable the development of spatial population distribution perspective models, improving our understanding of past and future demographic trends and enabling the development of public policies, particularly in contexts where we lack data. Thus, understanding how well remote sensing-based population grids represent the spatial distribution of population in a country with reliable data is essential for their development in population studies. We aim to detail an empirical analysis on the main top-down remote sensing-based population datasets over several years and with different spatial resolutions. Using Sweden’s highly reliable register-based grid data as a reference of the spatial distribution of its population, we aim to identify how close the main remotely sensed population datasets are from the reference and quantify this distance.

Keywords: Data and Methods, Remote sensing, Spatial Demography, Geographic Information Systems (GIS)

No extended abstract or paper available

  Presented in Session 173. Spatial Demography