LandScan

Abstract

Originally conceived for use by U.S. military and intelligence communities, the LandScan web application and population datasets, produced by East View Geospatial with data from Oak Ridge National Laboratory, are useful for various reasons including: spatial data analysis, humanitarian aid and relief, disease modeling, market growth, and sustainable development/environmental protection. Older datasets, back to 2000, are archived as new versions are made available. There are three main audiences of LandScan: researchers who download and use the datasets; researchers who use the WMS and WCS services; and those who use the web platform. This review focuses mainly on the web app.Search functionalities include the ability to drill down to the pixel level, i.e., one square kilometer grid units; to quickly get population counts for an area on the world map by pointing and clicking; and to get population counts for an area using the polygon drawing function.

Urbanizing the floodplain: global changes of imperviousness in flood-prone areas

“…Due to the variability of the available population datasets and the potential impact on estimates of population exposed to flood hazard [41], we used multiple datasets that satisfied the requirements of global coverage, spatial resolution less than 1 km, and multiple epochs. These datasets included: the 9 arc-second (∼250 m) GHSL dataset that represents population during 1990, 2000, and 2015 [42]; the Global Rural Urban Mapping Project that has 30 arc-second (∼1 km) resolution and represents 1990 and 2000 [43]; the Worldpop dataset that provides annual data from 2000 to 2020 on global population at a 3 arc-second spatial resolution [44]; the Gridded Population of the World dataset that represents population at 30 arc-seconds resolution every 5 years since 2000 [45]; and Landscan TM that provides annual population data since 2000 at 30 arc-seconds resolution globally [46]. Due to the scale discrepancy between the derived floodplain encroachment data (3 arc-seconds) and all the population datasets (9 and 30 arc-seconds) with the exception of Worldpop, we calculated the fractional areas of floodplain encroachment per flood probability as the number of 3 arc-second floodplain pixels for each return period T (masked to retain only pixels that were urbanized) divided by the number of non-null pixels within each 9 or 30 arc-second pixel.…”

Urbanizing the floodplain: Global changes of imperviousness in flood-prone areas

“…Due to the variability of the available population datasets and the potential impact on estimates of population exposed to flood risk 22 , we used multiple datasets that satisfied the requirements of global coverage, spatial resolution less than 1 km, and multiple epochs. These datasets included: the 9 arc-second (~250m) Global Human Settlement (GHSL) dataset that represents population during 1990, 2000, and 2015 49 ; the Global Rural Urban Mapping Project (GRUMP) that has 30 arcsecond (~1km) resolution and represents 1990 and 2000 50 ; the Worldpop dataset that provides annual data from 2000-2020 on global population at a 3 arc-second spatial resolution 51 ; the Gridded Population of the World (GPW) dataset that represents population at 30 arc-seconds resolution every 5 years since 2000 52 ; and Landscan™that provides annual population data since 2000 at 30 arc-seconds resolution globally 53 . Each dataset is adjusted to match United Nations Population Division (UNDP) national population estimates, but derived with a range of approaches that combine dasymetric algorithms and spatial data (e.g., nighttime lights, built-up area).…”