As mapping gains ground, a need to manage expectations

Immediately after the militant group Boko Haram attacked areas surrounding the fishing town of Baga in Nigeria early this month, media outlets were accused of turning a blind eye to the crime — a sobering reflection of the challenges of reporting from northern Nigeria amid conflict, the lamentable lack of reliable access to communications on the ground and, for many, the sad state of what is deemed newsworthy nowadays. Meanwhile, satellite images obtained by Human Rights Watch captured a clearer, albeit far from conclusive, picture of what happened: large-scale destruction in Doro Gowon, where the base of the Multinational Joint Task Force, a group that Nigeria, Chad and Niger started in 1998 initially to monitor crime within borders, and eventually, to conduct counterterrorism operations against Boko Haram, was located. HRW started reviewing and analyzing the raw images it acquired from French government space agency CNES and satellite service provider Astrium — with which it has a commercial agreement — in 2012, the same year that Josh Lyons, its satellite imagery analyst, joined the group. While HRW had already been turning to satellite images to detect human rights abuses since the early 2000s, much of the analytical work of these images was done externally — a practice that proved impractical given the nuanced nature of scrutinizing satellite imagery. “It can be very complex to interpret an exact sequence of events on the ground from just the images alone,” Lyons told Devex. “A very good analyst can come up with very good interpretations, but ultimately not be able to distinguish and choose between those two different and competing interpretations without some additional ground verification.” Satellite images as supporting information Conflict is becoming a more visible venue for mapping efforts, but deploying satellite technology in war-affected areas has not exactly been a recent development. For instance, in 2000, The New York Times turned to Ikonos, the first commercial high-resolution satellite, to depict before-and-after images of downtown Grozny, Chechnya, which had been flattened after Russia captured the city. While academic Lara Nettelfield pointed out that the Grozny photos “failed to arouse public sympathy or outrage for the plight of civilians in Chechnya,” the United Nations and nongovernmental organizations like HRW saw “commercial satellite technology … [as] definitely something that could be quite useful to complement the traditional forms of investigation on the ground,” according to Lyons. Through HRW’s network, Lyons coordinates with more than 100 researchers in different settings to help him interpret the raw images he sifts through. In certain cases, satellite imagery can verify eyewitness testimony that HRW has already collected. “The eyewitness, for example, will describe when their village was attacked, or when the military forces occupied a certain area. Or they will describe when there was a major offensive that caused thousands of people to move to a different part of the district, for example,” Lyons said. “So we’ll use the imagery to essentially validate and establish evidence that matches and supports what they’ve told us.” When on-the-ground information isn’t available or when it can’t send researchers to a certain site, HRW turns to social media for confirmation. A video uploaded by local citizens, activists, rebels and other armed forces or perhaps the military themselves, for example, can be matched against a satellite image to authenticate the analysis. In other instances, cross-referencing works the opposite way — that is, HRW uses an image to substantiate a video that is of primary interest. The analysis gleaned from the image and video then becomes the basis of a press release or a report that could be relevant to a human rights investigation. “The corroboration process is quite dynamic and works in different directions. It has to be flexible and pragmatic because every circumstance is potentially quite different,” Lyons said. “We need to have a very, very broad and multifaceted approach to how we collect as much and different types of information to help establish a higher degree of confidence and authenticity of exactly what has occurred on the ground and to validate the evidence that we’re going to build a case on.” Estimating the death toll, on the other hand, is a far more complicated process. Simply inferring deaths from the destruction of buildings is “extremely problematic” and even “essentially impossible,” according to Lyons. Some buildings, for instance, might have been uninhabited at the time of the attack. In rare cases, deaths can be identified through sightings of mass and individual graves. One of these exceptional instances is Syria, where HRW was able to estimate the number of people buried in a certain town in a given period based on the expansion of the cemetery, Lyons said. “Normally, in a major war, people simply don’t have the luxury or the opportunity to bury their family members in a normal cemetery,” Lyons said. “It’s too dangerous; they don’t have access. [Assessing the death toll] is not something we’d normally expect we’d be able to do with the imagery alone.” These difficulties make it unlikely for satellite images to serve as a stand-alone source for evaluating what is happening on the ground. “In very exceptional circumstances, satellite imagery might be one of many different inputs to a calculation, but it will never be a primary one,” Lyons said. Mapping as a collaborative process Despite the strides that satellite imagery has made in conflict-related work, the use of satellite images has been “more limited” in these settings, Lyons said. Satellite data for humanitarian operations, such as a disaster response, has been “more diverse and robust.” This was evident in the recent disaster response to Typhoon Haiyan in the Philippines, where the American Red Cross, for example, used open-source maps for relief distribution. Missing Maps, an initiative involving the American Red Cross, the British Red Cross, the Humanitarian OpenStreetMap Team and Médecins Sans Frontières that started out as an effort to map Mading, South Sudan, even looks beyond current crises by inputting satellite and on-the-ground data to OpenStreetMap and mapping previously “unmapped” areas that could be hotbeds for potential emergencies and outbreaks. The goal is “to put 20 million people on the map in two years” — an ambitious target that the Missing Maps team believes it can achieve, Dale Kunce, senior geospatial engineer and geographic information system team lead at the American Red Cross, told Devex. “In just the past three months, almost 1,000 people have made 1.7 million edits to OpenStreetMap in support of Missing Maps and the member organizations, literally putting entire cities on the map,” he said. But the project isn’t about imposing these maps on remote areas from afar. Through satellite mapping and on-the-ground field mapping with the communities concerned, Missing Maps, according to Kunce, allows locals to use the data “to advocate for themselves.” “The goal is to work with communities to create the map of their​ community,” he said. “We do not own the map or data and are simply providing some technical expertise to make mapping their community a little easier.” ‘Do no harm’ as an operating principle While InterAction, a network of international NGOs based in the United States, has used its members’ voluntarily provided information instead of satellite data in its own mapping projects, it has also worked with a similar principle. Soon after the 2010 Haiti earthquake, InterAction put together the Haiti Aid Map, which gave their members as well as any interested donors or organizations an idea of what kind of work was being done in the wake of the disaster. It was also the first NGO Aid Map. According to Julie Montgomery, director for innovation and learning at InterAction, the organization faced three challenges: getting accurate information in a rapidly changing environment, ensuring that it wasn’t duplicating the data collection efforts of other organizations such as the U.N. Office for the Coordination of Humanitarian Affairs, and preventing the safety of aid workers from being jeopardized. “While we are strong advocates of openness, we realize there can be negative consequences to sharing data. One of our five operating principles is ‘do no harm,’” Montgomery told Devex. “We let organizations know that they should only share information if it is safe to do so, and review data that has been submitted to make sure that there isn’t anything that could put anyone in harm’s way.” In countries where restrictive laws limit the ability of NGOs to operate, InterAction recommends its members to exercise the same discretion. For instance, during a review of NGOs receiving foreign funding in China, the organization asked its members whether changes had to be made in the data they had provided about their work in the country. The Pakistan earthquake in 2013 triggered a similarly cautious move. “We made a deliberate decision to not map our members’ work because of security reasons,” Montgomery said. In some cases, removing specific details — the names of local partners and the size of the budget, for example — may be deemed sufficient to protect these groups. But while these precautionary measures have protected InterAction members, NGO Aid Map wouldn’t have taken off as a tool for awareness and coordination if InterAction members had kept mum about their projects. To paint an accurate portrait of activities, InterAction had to understand its members’ motivations for sharing their data and reassure them that the effort will be worthwhile. Today, NGO Aid Map covers the work of all InterAction members in all countries and sectors — no small feat for organizations that gather and store data in different ways. “We also needed to make it easy, or as easy as possible, for members to share data,” Montgomery said. Mapping as an investment Some organizations may understandably have qualms about using mapping tools — whether in the form of satellite imagery or a more simple method, such as InterAction’s — in their operations. “They think using satellite imagery is extremely complex, far too expensive and far too hard, and I think it’s a false impression,” Lyons said. “There are definitely opportunities for people to bring that expertise to their group.” There’s also the issue of the expenses associated with acquiring and analyzing these images. But Lyons argued that compared with the cost of sending staff to faraway locations, spending on boosting the research capacity for satellite data and imagery could prove to be a smaller investment. “Sitting at my desk and looking at an image can be quite cheap, in that context,” he said. “I think that many NGOs spend far more on phone calls than we do on satellite imagery.” Donors and foundations, he added, are always looking to support new technology adoption by NGOs — “and that’s exactly been the case for HRW.” For InterAction’s Haiti Aid Map, a project largely funded by FedEx and supported by the U.S. Chamber of Commerce’s Business Civic Leadership Center, such partnerships have helped keep the operation going. Still, for mapping efforts like Missing Maps, the money needed to sustain the project has not yet flowed as easily. “We haven’t yet secured the top-level donors that we need to ensure the project’s success,” Kunce said. The Missing Maps team, according to Kunce, is currently looking to work with corporate and government partners who can supply the necessary technology, funding and data for the initiative. “The on-the-ground piece is expensive, but the data and the enhanced community participation and capacity are well worth it,” he said. How else can development and humanitarian organizations improve the way they use mapping technologies in guiding their work? Let us know by leaving a comment below. Check out more insights and analysis provided to hundreds of Executive Members worldwide, and subscribe to the Development Insider to receive the latest news, trends and policies that influence your organization.