Human geography researchers use residential census data and open-source information to create models of populations across the globe. However, census data for some parts of the world are sparse. To provide a more comprehensive view of the world and its inhabitants, researchers at the Department of Energy’s Oak Ridge National Laboratory are turning to satellite images of homes under construction.
"We have been struggling to find data on dwelling size in Sub-Saharan Africa and other data-sparse areas for years," said ORNL’s Carson Woody, a human dynamics researcher. “In rapidly developing countries, imagery of construction provides information to fill in the gaps.” Woody coined the term “waffle home” to describe what the houses look like from space while being built, which includes repeating patterns of raised squares or rectangles that resemble waffles. Her work was published in the proceedings of the GIScience conference in Leeds, UK.
Woody works on Population Density Tables, a machine learning method to report building occupancy at regional and national levels for more than 60 building types, such as residential or commercial. As people in developing urban regions of the world move into different types of dwellings, systems such as Population Density Tables must be updated to reflect changes. While looking at developing parts of the world where census data about residential house size isn’t as plentiful, Woody noticed that houses in the process of being built offer information that can be converted into the type of data she needed.
Woody leaned on her colleague Tyler Frazier, an R&D associate at ORNL with project and field research experience in Ghana from 2007 to 2013. Using high-resolution satellite imagery over Ghana, they viewed homes that had walls built but lacked a roof. By digitally measuring the waffle home rooms, they could calculate the size of the house and, using commonly found information on average household size, determine how many people likely will reside in the completed house.
Woody’s next step is to expand this method to other countries with a similar data gap in residential populations. Additionally, her work supports ORNL’s national security mission by providing better estimates of where people may be during the day. This information helps improve emergency response for conflicts and natural disasters.
UT-Battelle manages ORNL for the Department of Energy’s Office of Science, the single largest supporter of basic research in the physical sciences in the United States. The Office of Science is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science.