Improving the accuracy of gridded population estimates in cities and slums to monitor SDG 11: Evidence from a simulation study in Namibia

People living in slums and other deprived areas low- middle-income country (LMIC) cities are under-represented censuses, subsequently "top-down" census-derived gridded population estimates. Modelled data a unique source of disaggregated information to calculate local development indicators such as the Sustainable Development Goals (SDGs). This study evaluates if, how, “top-down” WorldPop Global (WPG) Unconstrained Constrained datasets might be improved simulated LMIC urban by incorporating slum counts into model training. We found that WPG-Unconstrained model, with or without training data, underestimated while overestimating rural areas. The percent (SDG 11.1.1), for example, was estimated 20% less compared "true" value 29.5%. WPG-Constrained which included building footprint auxiliary datasets, far more accurately all grid cells (including areas), inclusion further estimates SDG 11.1.1 at 27.1% 27.0%, respectively. Inclusion metrics models can substantially improve cell-level accuracy both • Slum area populations censuses. Gridded derived from census inherit inaccuracies. inputs exacerbate inaccuracies Accuracy is greatly data. Settlement typology layers accuracy.