Understanding where people are located, and how those locations change between day and night, is essential for effective planning and decision-making. LandScan Mosaic is a high-resolution population modeling framework developed at Oak Ridge National Laboratory that uses buildings, land use, points of interest, remote sensing, and other geospatial data to estimate daytime, nighttime, and 24-hour ambient population distributions. The framework also reports uncertainty, helping users distinguish between locations supported by strong evidence and those where estimates are less certain.

This presentation will introduce the scientific approach behind LandScan Mosaic and demonstrate how high-resolution population estimates can support disaster response, infrastructure and transportation planning, public health, exposure and risk assessment, and humanitarian applications. It will also discuss how building-level modeling provides a more realistic representation of human presence than conventional population maps, particularly where the timing and precise location of populations matter.