LandScan Looks to the Future
Satellite imagery has the potential for a range of uses
When tsunamis struck the Indonesian coast in December 2004, the destruction left relief agencies scrambling to locate thousands of stranded people in need of assistance. LandScan, an ORNL-developed global population database that shows geographical distribution of population at 1-kilometer resolution, was a crucial part of the response. Relief workers used LandScan population distribution maps to quickly determine the locations and numbers of potential tsunami victims who would have otherwise been cut off from communication.
LandScan can combine satelliteb imagery with a range of other information sources, such as census data, as shown in this graphical representation of San Diego's daytime population.
LandScan, first developed at ORNL in the late 1990s, has since grown into the community standard for mapping global population distribution. For example, LandScan provides the data necessary for siting reactors away from heavily populated areas as well as providing baseline information for disaster response, humanitarian relief, sustainable development, environmental protection and national security.
By integrating the best available census and remote-sensing data into a geographic information system framework, LandScan Global describes the total population for a one-square-kilometer area over an average 24-hour period. For the United States, LandScan USA provides even finer resolution, measuring population at the scale of a single city block for nighttime as well as daytime scenarios. In the wake of Hurricane Katrina in 2005, for example, LandScan-produced images were used to brief President Bush about the coast's affected population.
Now, Budhendra Bhaduri, a principal member of the LandScan team, says the award-winning population database is moving into new territory. "The emphasis of LandScan has always been on how many people are where," says Bhaduri, who leads ORNL's Geographic Information Science and Technology group. "We are turning the focus to ask not just how many people are there, but who are these people?"
The answer to this question lies in what Bhaduri calls the "geovisualization of the invisible." By analyzing satellite images, ORNL researchers hope to use geographic indicators such as the structure of neighborhoods, location of marketplaces or construction of satellite towers to understand how people live.
Geographic patterns could reveal, for instance, the economic strength of a given population, which could then be linked to other socioeconomic variables such as level of education, access to information, average size of family or access to services and facilities. Information gleaned from satellite imagery could be used to remotely monitor situations in regions where ground access might be limited. This type of geographic analysis coupled with LandScan's capabilities could be used, for example, to track the living conditions of citizens in Afghanistan and Iraq as U.S. troops withdraw.
"As we start pulling out of Afghanistan, how do we monitor that infrastructure and ensure that things are improving?" Bhaduri says. "That's when we start to look for the sort of indicators we can understand. Are more higher-income neighborhoods, roads, satellite towers, electricity transmission towers and lines being built? Are markets growing or shrinking? All these factors are indicators of stability and prosperity. It all comes down to understanding the patterns on the imagery."—Morgan McCorkle