Using an innovative approach that combines geospatial science, remote sensing technology, and machine learning algorithms, LandScan Global is the finest resolution global population distribution data available representing an ambient (24 hour average) population. The LandScan Global algorithm, an R&D 100 Award Winner, uses spatial data, high-resolution imagery exploitation, and a multi-variable dasymetric modeling approach to disaggregate census counts within an administrative boundary. Since no single population distribution model can account for the differences in spatial data availability, quality, scale, and accuracy as well as the differences in cultural settlement practices, LandScan population distribution models are tailored to match the data conditions and geographical nature of each individual country and region. By modeling an ambient population, LandScan Global captures the full potential activity space of people throughout the course of the day and night rather than just a residential location.
