Satellite Human Rights and the importance of imagery

When my friend Janet Haven came to visit a few weeks ago, she came armed with a pair of images. Produced by mapping experts at the American Association for the Advancement of Science, they show Porta Farm, a large settlement in Zimbabwe just outside Harare in 2002 and again in 2006. In the four years between the image, thousands of structures, probably housing 6-10 thousand people have disappeared. They’re a powerful demonstration of the relocations of informal settlements implemented by the Mugabe government – Operation Murambatsvina. UN reports estimate than between 700,000 and 2.4 million Zimbabweans were moved from their homes as part of the operation, which many believe was designed to punish Mugabe’s political opponents. They’re also a powerful example of the impact geographic imagery can have on human rights cases.

I figured these were the sort of amazing pictures that you occasionally get to see in the course of doing the technology/activism/globalism work I do, but never get to blog. But the AAAS put out a press release yesterday – blogged by Jason Kottke – and are now flying around the web… which encouraged me to take a closer look at how one finds and analyzes images like these.

The source images for this analysis come from Digital Globe, a provider of satellite imagery to everyone from Google Earth to military and industrial users. Sources in Zimbabwe were able to give AAAS researchers accurate latitude and longitude coordinates for the farm – AAAS purchased high quality images from Digital Globe for the area near the coordinates, choosing images taken with a minimum of cloud cover, then used image processing techniques to identify the centers of structures, and statistical estimation techniques to estimate the population of the settlement.

The Digital Globe site isn’t nearly as easy to use as Google Earth (I get the sense their server may currently be swamped), but it provides imagery of parts of the world I’m interested in at much higher resolution than Google Earth does. Whether or not the costs of purchasing imagery like that shown by AAAS is prohibitively expensive for the amateur researcher, I don’t know – I’m currently trying and failing to purchase some images of Accra. But it’s fascinating to see the power of these images and to think about some of the changes one could document with comparative satellite imagery.