Satellite Imagery Can Now Be Used To Predict Poverty Regions Better Than Ever

Science Jeanette Walsh
http://stocknewsusa.com/wp-content/uploads/2016/08/Satellite-Imagery-Can-Now-Be-Used-To-Predict-Poverty-Regions-Better-Than-Ever-1024x512.jpg

Research from the Journal Science has shown how satellite imagery and algorithms can be used to predict poverty in regions where such data would otherwise be unavailable. The non-availability of constant and reliable data in developing countries is one mitigating factor to address the problem.

Researchers from Stanford University-trained computer systems to identify many impoverished areas from survey and satellite data across five African nations.

The research team, which consisted of Marshall Burke, Neal Jean and colleagues said the technique could be used to monitor and address poverty in developing countries across the world.

According to the World Bank, anyone living on less than $1 a day is considered poor. In an interview with BBC’s Science in Action initiative, Dr. Burke, a Sanford assistant professor of Earth system science stated that his team collects poverty data through household surveys conducted by enumerators who move about different houses asking questions about consumption, income and what was earned the year before. The data is the used to generate their poverty measures.

Night Lights

The surveys are however infrequent, costly and at times almost impossible to conduct in certain regions of countries because of issues such as armed conflicts or political instability.

For this reason, there’s need to develop other more accurate means of measuring income and household consumption in developing countries.

Mapping poverty through satellite imagery isn’t something new. Studies have indicated that space-based data that captures night lights can also be used to predict wealth in a particular area.

However, night lights don’t pose such a good indicator for lower end income distribution, where images are dark.

The study focused on images captured during daylight with features like metal roofs, paved roads, waterways – indicators that can distinguish the different economics in developing countries. The model was used to categorize daytime images of Tanzania, Malawi, Rwanda, Uganda and Nigeria.