LandScan USA—Scanning High and Low, Day and Night

Since the late 1990s, Oak Ridge National Laboratory has developed the world’s finest population distribution model and data, called LandScan, which allows us to understand how people choose to inhabit space and time. The model is resolved at 30 arc seconds (approximately 1 kilometer) resolution globally (LandScan Global) and at 3 arc seconds (approximately 90 meters) in the United States (LandScan USA), where it was the first to consistently model daytime population distribution compared to nighttime residential population that is typically reflected in the census. LandScan USA was used to visualize the activity-based dispersion of nighttime residential population during the daytime around the city of San Diego, California. While nighttime population is primarily restricted to residential areas and the suburbs throughout the diverse terrain, daytime population tends to be heavily concentrated in downtown San Diego, other commercial areas, shopping malls, schools, colleges, and universities, illustrating the movement of people out of the suburbs and residential areas.

LandScan is a dasymetric geospatial model that decomposes best available census data to finer space/time resolutions using geographic, physiographic, socioeconomic, and cultural indicator data and high-resolution imagery. The ancillary indicator data used to distribute the nighttime population includes census data at the block level, land cover, land use, slope, and distance from various types of roads. Locating daytime populations requires not only census data but also other socioeconomic data including places of work; journey to work; additional mobility factors such as daytime business and cultural attractions; and populated places datasets including prisons, schools, and other academic institutions. Verification and validation of the model are accomplished in two ways: input data to the model is extensively quality assured to account for inaccurate, missing, or outdated input variables, and model output is compared to high-resolution imagery to ensure the spatial accuracy of population distribution.

LandScan and LandScan USA have greatly extended capability in geospatial modeling, analysis, and visualization in missions of international significance such as disaster response, humanitarian relief, socioenvironmental studies, and homeland and national security.

Produced by UT-Battelle, LLC, under Contract No. DE-AC05-00OR22725 with the Department of Energy