Eyes in the sky: Remote sensing in the service of human rights

If scientific endeavor has a purpose beyond the accumulation of knowledge for its own sake, it must be for the progress and betterment of humanity. Despite the incredible scientific developments that have defined the modern era, however, the ability of science to solve humankind's most enduring problems remains elusive. Nowhere is that more apparent than in conflict-torn regions where weapons technologies, with their roots in scientific discovery, are exploited with lethal effectiveness.

While technology's power to exacerbate violence in that way is well established, its role in mitigating hostilities has often been considerably less visible. Recently, however, new natural science technologies have made it possible to take a more active role in conflict reduction and the promotion of universal human rights. One of the most powerful of those technologies is satellite imagery.

Once the exclusive purview of cold war intelligence agencies, satellite imagery has, in the past decade, become an indispensable tool in civilian applications ranging from navigation to crop monitoring. For those of us working in the human-rights field, it has provided a new window into areas of the world once off-limits for reasons of politics, security, or physical inaccessibility, and where abuses were known only through the erratic descriptions of those who managed to escape.

Captured by meter-sized telescopes in low, Sun-synchronous orbits around Earth, civilian satellite imagery is available at spatial resolutions as high as 50 centimeters per pixel, and in wavelengths ranging from the visible to L-band synthetic aperture radar. Because it can rapidly and quantitatively document human rights violations almost anywhere on Earth, satellite imagery has revolutionized the field.

For all the capabilities that these tools offer, however, they are of little use until we know where to point them. Accurate field reports are critical to image targeting and acquisition, which is why at the American Association for the Advancement of Science (AAAS) Geospatial Technologies and Human Rights Project we partner with a broad range of human-rights organizations operating worldwide, from household names to smaller groups that focus on a particular country or region.

Even so, associating accounts of human-rights violations with a precise set of geographic coordinates presents a challenging endeavor. In remote areas of Africa and Southeast Asia, for example, up-to-date maps are often unavailable, and many existing settlements simply do not appear on them. Even when accurate surveys exist, separate groups will often use different names to refer to the same location, and those discrepancies are seldom reported. Transliteration introduces yet another layer of complication. Depending upon phonetic convention, a single place name could have dozens of variant spellings in the Latin alphabet.

To combat that problem, we employ a number of techniques, ranging from 'fuzzy' matching on database queries and analysis of low-resolution satellite imagery, to the simple use of distances, bearings, and landmarks to pinpoint locations for the often-fragmentary reports arriving from the field. It can be slow, frustrating work, but when it succeeds, it reveals to the world actions that perpetrators would prefer to keep hidden.

This past summer, for example, we partnered with Amnesty International to analyze the effects of the escalating Syrian civil war on the city of Aleppo. By comparing images covering nearly 200 square kilometers of the city over a two-week period in August, we were able to identify signs of conflict, including military activity, property destruction, shell craters, and the erection of more than 100 roadblocks around the city. We also found numerous destroyed and damaged buildings; in one instance, a tank was observed next to a structure that was later demolished. At the airport, two helicopters with dimensions and rotor configurations matching Mi-24 'Hind' attack helicopters had changed locations, suggesting recent use, while elsewhere dozens of heavy trucks and vehicle trailers had been removed from a military depot.

Similar phenomena were visible during the conflict in Libya in 2011, where tanks were used for heavy fighting in the city of Misurata. Although the camouflage paint of those vehicles rendered them difficult to distinguish from vegetation in the visible-band imagery, when near-IR wavelengths were employed they became easily identifiable.

Indirect techniques can also be used to gather information about a conflict. During the 2009 conflict in Sri Lanka, for example, shell craters were spotted in a crowded area that had been designated as a 'civilian safe zone.' By exploiting the bilateral symmetry of these features, we were able to compute the azimuths of the incoming projectiles, and then trace them back to combatants' mortar positions. In that same series of images, we could also watch graveyards expand over a period of mere days.

Remote sensing is being used not only for rapid response to developing crises, but also for long-term monitoring. As part of the Eyes on Nigeria project, for example, we analyzed a number of human rights abuses throughout that country over an extended time period. In the city of Port Harcourt, we used high-resolution imagery from 2008 and 2010 to identify nearly 400 structures that had been removed as part of a slum-clearing program. We were able to confirm reports elsewhere in the Niger River delta that civilians had been harmed during combat between government forces and rebel groups.

A similar analysis was carried out in the city of Jos in Plateau State where, in 2010, violence flared between Muslims and Christians. Results indicated that nearly a square kilometer's worth of total damage had taken place at multiple locations throughout the city. In response to reports of multiple oil spills near the Niger Delta town of Bodo, we obtained imagery of the area covering a five-year interval from 2006 to 2011. By analyzing the terrain's near-IR reflectivity, we determined that an area covering three square kilometers around the settlement had suffered almost complete vegetation death.

Whereas a single town can be analyzed using high-resolution imagery, analysis on a regional scale requires a different approach. For example, to document the practice of industrial gas flaring, we relied upon a database developed by volcanologists at the University of Hawaii to detect hotspots in low-resolution thermal IR imagery collected by NASA's MODIS instruments. By obtaining all such detections in the Niger Delta from 2000 to 2011, we could identify persistent clusters of abnormally high temperatures, calibrate them using ground-truth measurements collected by local researchers, and obtain high-resolution imagery for select areas to determine the effect the flaring has had on nearby settlements.

Buoyed by the success of our work in Nigeria, AAAS is currently expanding our research in the domain of environmental justice. Our goal is to find even more human rights applications for environmental management techniques. Right now, for instance, we are employing a decade's worth of MODIS data to identify sources of water pollution in Asia. We are also exploring methods to quantify land under shifting cultivation, the practitioners of which are often vulnerable populations subject to forced relocations as countries build modern infrastructure. If successful, that research could enable those groups to resist relocation or advocate for fair compensation for their land.

Scientific research has ever been a powerful force in human progress. For the two of us, the opportunity to integrate the fascination and excitement of scientific endeavor with the fulfillment of social justice has been an irresistible draw. That spirit is reflected in the AAAS mission statement, which calls on us 'to advance science, engineering, and innovation throughout the world for the benefit of all people.' In partnership with human-rights groups, academics, and individuals worldwide, we intend to do exactly that by continuing the development of new applications of space-based remote sensing to help forge a better future.

Jonathan Drake is the senior imagery analyst at the AAAS's Geospatial Technologies and Human Rights Project. He holds a bachelor's degree in physics from Dickinson College and a masters from the school of Earth and space exploration at Arizona State University, where he acquired his knowledge of remote sensing by working with the THEMIS instrument aboard the Mars Odyssey spacecraft. At AAAS, he has been involved in developing applications of radar and thermal imaging to problems relevant to human rights and in analyzing visible and near-IR imagery in support of program objectives.

Eric Ashcroft is a geospatial analyst at the AAAS Geospatial Technologies and Human Rights Project and has a master's degree in geography from the George Washington University. While there, he worked as a research assistant mapping health and wealth in Accra, Ghana using geographical information systems and remotely sensed imagery. His other work includes using remotely sensed data to monitor riparian health in Victoria, Australia.