Satellite data used to pinpoint the poorest areas of the world

BBC : Identifying the poorest of Earth’s residents can be difficult for various reasons, including their remoteness or local political instabilities. Although satellites have been used to map poverty by looking for the most sparsely illuminated areas, that criterion has not proven definitive. Now a team from Stanford University has combined those nighttime maps with high-resolution daylight images to look for indications of different levels of economic well-being, such as the number of paved roads compared with unpaved ones and the presence of metal roofs on buildings. Sophisticated computer software then categorizes the various indicators and looks for details and patterns that are “predictive of poverty,” says Marshall Burke, one of the team members. To verify the computer model’s accuracy, the researchers compared the results with household survey data.