Program in GIS

Global Ambient Population, 2002

Source:  Oakridge National Laboratory, http://www.ornl.gov/sci/gist/landscan/index.html.
 

Describe the LandScan global population dataset. This dataset, which is produced by the Oakridge National Laboratory and managed by UT-Battelle for the Department of Energy, provides worldwide coverage of the earth's "ambient" population (the average population for a given location over 24 hours) on a 1 kilometer square grid.  According to the LandScan Global Population Databases Fact Sheet, LandScan:

• Is the world's most accurate and reliable, geographically based, time-of-day population distribution model and databases
• Allows quick and easy assessment, estimation, and visualization of population at risk
• Provides high resolution nighttime (residential) as well as daytime population distributions
• Is a critical component of emergency planning and management, rapid risk assessment, evacuation planning, consequence assessment, mitigation planning and implementation.

The data can be viewed and analyzed using the ESRI ArcView product as the files are stored in the ESRI grid format.  The data are copyrighted and the permission of Oakridge National Laboratories is required to access and use the data for both commercial and non-commercial purposes.

What is its spatial resolution? The data are compiled on a 30" by 30" latitude/longitude grid.

What are the dimensions of the land areas that correspond to each grid cell? The dimension of the area encompassed by a 30" by 30" latitude/longitude grid varies by latitude.  The grid is 1 kilometer by 1 kilometer, which is 1 square kilometer.

What does the term "ambient population" mean? The ambient population, as opposed to residential population, takes into account the movements of individuals through a given area (or cell), thus giving a geographically based, averaged over 24 hours, "time-of-day" population.  For instance, no individuals live on an interstate, but they do travel on it from time to time.  A resident population would show no one living on the interstate, whereas the ambient population would indicate the "presence" of individuals (a population, if you will) based on factors specific to the interstate (see Input Variables below).

What are the data used for? The data are used by various agencies of the government to assess "at risk" populations should an area be hit by a disaster, natural or man made, of some type—at risk indicating susceptible individuals.  Knowing where these "at risk" populations are located, governments can make arrangements for emergency planning, resources, and response.

Input Variables

Identify the input variables used to estimate the ambient population of each grid cell. The input variables include roads, slope, land cover, populated places, nighttime lights, exclusion areas, urban density factor, and coastlines.

Explain how each input variable contributes to the estimate. The availability and nature of input variables impact the probability coefficient calculations used to determine ambient population for each cell.

• Roads—these are the number one indicator of population.  The density of roads is suggestive of population as they are vital to human settlements.  Roads are weighted by distance from major roads.
• Slope—as slope increases, population density tends to decrease.  Thus, slope plays an important role in the calculation of the coefficient by noting the gradient for each 1 kilometer cell.  Slope categories are weighted by their favorability for population.
• Land Cover—this is a good indicator of population density (and, perhaps, the best one).  By assessing land cover types and determining the average density for the type, land cover can be used as a first pass assessment of population.  Land cover is weighted by type with exclusions for some types.
• Populated Places—these identify specific named or un-named populations of a particular location and size.  They can be used to check against the nighttime lights variable and allow a greater weight to be assigned to those places that are both populated and have nighttime lights.
• Nighttime Lights—the amount of light emanating from a particular cell is an indication of the presence of population, except where lights are not dependent upon people, such as the lights along an empty highway.  Lights, however, reach a saturation point where they are not able to indicate high density populations and sometimes, at the low end, do not show the absence of population.  Nighttime lights is weighted by frequency.
• Exclusion Areas—some cells, such as those with water and ice, have no or show no indicators or population.  Such cells, after comparing them with the roads, populated places, land cover, and nighttime lights information, are assigned a population value of zero.
• Urban Density Factor—when populated places and nighttime lights overlap, the variable is given a higher probability weighting than cells that show nighttime lights alone.
• Coastlines—checking the position of the data against coastline data helps to ensure that mismatches are corrected, since it is easy to see where data extend beyond a known coastal boundary

By using the variables to determine a probability coefficient, it is possible to impact census population counts to come up with an ambient population for a given cell.  As would be expected, coefficients vary widely from country to country and region to region.

Organizations Involved

Identify the organizations involved in creating and distributing the dataset. The dataset is organized and run by Oakridge National Laboratory's Global Population Project, in conjunction with UT-Battelle, and uses data from:

• U.S. Census Bureau
• National Center for Geographic Information and Analysis
• United Nations Statistical Division
• National Imagery and Mapping Agency (NIMA)
• U.S. Geological Survey (USGS)
• National Oceanic and Atmospheric Administration (NOAA)

How old is the most recent version created? The most recent version of the data set is "Global Ambient Population, 2002."

How frequently is the dataset revised?  The data set has been updated annually from 2000 to 2002, from an original dataset compiled in 1998.

Acquiring the LandScan Data

Where can you get the LandScan dataset?  Until March 2004, the LandScan Global Population Data was available online at http://www.public.ornl.gov/landscan/Nlandscan_main.cfm (a link is not provided as the site no longer exists, but is provided for reference).

Since the site for this has been removed by Oakridge National Laboratory, if you would like to use the dataset for either commercial or non-commercial uses, you should contact:

Margaret Spurlin via email at spurlinms@ornl.gov.

No reason has been given for the change in the distribution of the data set.

Oakridge National Laboratory LandScan 2000 Global Population Database Release.  http://www.ornl.gov/sci/gist/landscan/landscan_2000_release.htm. Accessed 10 March 2004.

LandScan Global Population Databases Fact Sheet, Oakridge National Laboratory.  http://computing.ornl.gov/cse_home/LandScan%20long.pdf. Accessed 28 February 2004.

Oakridge National Laboratory LandScan Global Population 1998 Database. http://www.ornl.gov/sci/gist/projects/LandScan/landscan_doc.htm. Accessed 19 February 2004.