Modeling Population Distribution Based on EO-Derived Data on the Built-Environment

Abstract

Last years, German Aerospace Center (DLR) made very big step forward in the context of global urban mapping. TanDEM-X mission, utilizing TerraSAR-X and TanDEM-X satellites, made it possible to derive very high resolution built-environment raster products – e.g., the Global Urban Footprint (GUF) settlement mask (Esch et al., 2013). The goal of this master thesis project is to investigate the potential to model the human population distributions based on a combination of the above mentioned Global Urban Footprint product, statistical census data and – optionally - additional land cover maps (e.g., CORINE Land Cover). Resulting layers on the spatial distribution of population – once provided on global level - would be highly beneficial for sustainable spatial and environmental planning, land management as well as policy. To conduct the study, algorithms were implemented as a tool for ArcGIS that were based on spatial/areal weighting and dasymetric mapping technics. In general, this includes the spatial disaggregating of information (Sleeter, 2004). By comparing with officially available numbers, the results of evaluation showed good enough accuracy for the test area, namely Federal State of Bavaria. Further modifications and developments of this project are possible.