World Settlement Footprint 3D - A first three-dimensional survey of the global building stock

Social cartography and satellite-derived building coverage for post-census population estimates in difficult-to-access regions of Colombia

“…The current model differed from previous examples because it included a sub-model that explicitly estimated building counts for inaccessible locations. This was necessary because we did not have enumerations of buildings from the census cartography nor from remote-sensing data; our remotely-sensed building data (Esch et al 2022) measured building coverage for each 90 m grid cell but did not include individual building footprints. Because of the hierarchical nature of the statistical model, the uncertainty around our population estimates also accounted for uncertainty in building estimates.…”

“…We used estimates of total building area per 90 m grid cell obtained from World Settlement Footprint 3D (Fig. 2; Esch et al 2020;Esch et al 2022). This provides essential information about where buildings are located in these remote areas and gives an indication of how many buildings are likely in each location.…”

“…The estimated building areas were validated using very high-resolution building models (<50 cm) available from 19 regions worldwide (Esch et al 2022). The accuracy assessments showed a slight bias towards overestimation globally with mean errors (ME) ranging from -6.48% to 12.99%.…”

Small area population denominators for improved disease surveillance and response

“…Moreover, the development of approaches for estimating building heights and volumes from satellites (e.g. Esch et al, 2022 ) presents opportunities to account for high rise residential or commercial buildings in estimation modelling. Geostatistical modelling from GPS-located survey data also offer solutions for the small area mapping of population demographics to move beyond large area summaries or outdated census data (e.g.…”

DataSheet1.pdf

“…The Copernicus Land Monitoring Service updates the CORINE Land Cover dataset (Büttner, 2014) providing 27 land cover classes at 6-year intervals but only provides coverage of Europe (39 countries). Some promising global morphology datasets derived from satellite data have begun to be reported (Esch et al, 2022), but as of writing are not yet publicly available.…”