Population density estimations for disaster management: Case study rural zimbabwe

Abstract

This paper tackles the need of enhanced population data for disaster management and aid delivery studies in developing countries. It analyses the usefulness of a set of spatial data layers, including medium resolution satellite imagery, for population density estimations in rural Zimbabwe. The exercise conducted on a 185 × 185km area at a grid cell size of 150m allowed us to develop a methodology that can be extended to the whole of Zimbabwe. The surface modelling of population density was implemented by integrating 4 main variables: land use, settlements, road network, and slopes. During the modelling procedure, pixel weighting values were allocated according to pre-defined decision rules. In a final step the district population counts of the recent Zimbabwean census were distributed among all pixels of the relevant district according to the pixel weighting values. The resulting land use information and population data can be linked to vulnerability and food insecurity. In order to be transferred to other countries, the modelling procedure needs to be adapted to case specific characteristics, the determination of which requires a certain level of local / expert knowledge. In addition, passive sensors might not provide sufficient cloud free satellite data for regions lying within the moist tropics. © 2005 Springer-Verlag Berlin Heidelberg.