Mapping 20 Years of Urban Expansion in 45 Urban Areas of Sub-Saharan Africa

By 2050, half of the net increase in world’s population is expected to reside sub-Saharan Africa (SSA), driving high urbanization rates and drastic land cover changes. However, data-scarce environment SSA limits our understanding urban dynamics region. In this context, Earth Observation (EO) an opportunity gather accurate up-to-date spatial information on extents. During last decade, adoption open-access policies by major EO programs (CBERS, Landsat, Sentinel) has allowed production several global resolution (10–30 m) maps human settlements. mapping accuracies are usually lower, limited lack reference datasets support training validation classification models. Here we propose a approach based multi-sensor satellite imagery (Landsat, Sentinel-1, Envisat, ERS) volunteered geographic (OpenStreetMap) solve challenges remote sensing SSA. The proposed assessed 17 case studies for average F1-score 0.93, applied 45 areas produce dataset expansion from 1995 2015. Across studies, built-up averaged compound annual growth rate 5.5% between comparison with local reveals heterogeneity Overall, densities decreasing. impact differs depending size area its income class.