A Random Forest Dasymetric Approach For Mapping The Population Distribution At High Spatial Resolution

recisely locating the population is a sustainable development challenge for both regional and local public authorities. In Wallonia, Belgium, the population density of the Walloon municipalities shows significant variability. Like all territories, Wallonia is changing, transformed by a set of spatial planning decisions and regional policies. Knowledge of these aspects is essential to understand and solve problems in many areas. A fine-scale and precise population distribution map, combined with high- resolution thematic data such as land use and land cover, is essential for analysing urban development, monitoring human-environment interactions and spatial planning. Very high-resolution maps are produced from National Register (RNPP) data but they are not freely accessible due to privacy or only in a degraded format. In this paper, we propose a Random Forest algorithm based dasymetric method for the development of an up-to-date and open data population density map at 100m resolution over Wallonia.
info:eu-repo/semantics/published