Mapping Human Terrain

If you walk the hallways--that is, if you're escorted through the hallways--of the National Geospatial-Intelligence Agency in Bethesda, Md., you'll see some remarkable maps.

Based on photos taken from space, they depict cities, highways, farm fields, forests and rivers in exquisite detail and vivid color. But these are mere decorations compared with the highly detailed and highly classified images that geospatial analysts here study daily behind heavily bolted doors.

Working with satellite images, video shot by unmanned aerial vehicles and data collected from spy planes using a variety of other overhead sensors, NGA provides the U.S. armed services with maps, images and targeting analysis to guide military operations.

Agency Director Letitia Long provided rare insight into NGA on May 2 when she issued a statement highlighting its role in the discovery and killing of Osama bin Laden.

"NGA applied a range of geospatial- intelligence capabilities, including imagery, geospatial and targeting analysis, along with image sciences and modeling" that helped guide the U.S. special forces who located and killed bin Laden in Pakistan on May 1, she said.

Publicly discussing NGA's role in a risky military operation was uncharacteristic of the secretive intelligence agency. But Long, who has been NGA director since August 2010, has been pushing the agency to move beyond its comfortable stock-in-trade of maps, photos and other images. While those products are critical for revealing what's happening and where and when it's happening, and even who might be involved, Long has told her agency that it also should be trying to predict what will happen.

NGA, which is relocating employees to its new headquarters building in Springfield, Va., should "be moving to more of an anticipatory posture," she said in an address to a geospatial intelligence symposium in November 2010.

Rather than report what is happening, NGA analysis should aim to predict "what could happen, where it could happen and why it could happen," Long said. That way "we can create new value for the policymakers, the warfighters, the intelligence community and first responders," she said.

The key to predicting the future? "Human geography," Long says.

Maps, photos and other images provide information about the tangible features of a particular location--the buildings, the roads, the mountains and rivers--that's physical geography. And it's valuable, as far as it goes.

But the human geography of the same place would reveal much more--tribal boundaries, political ideology, ethnicity and languages. According to Long, it could include such elements as birth and death rates, degree of education, access to media, principal market commodities and proximity to health care facilities.

All are factors that would influence human behavior.

Unlike natural terrain and man-made features, the human geography "data set can and does change rapidly and dramatically based on the problem," Long said. Geospatial intelligence "is the examination of all this data viewed through a spatial and temporal lens."

That is space and time--where and when events happen.

Long contends that human geography "can yield new insight" that would help answer such questions as:

-- Where are conditions right for weapons of mass destruction

proliferation?

-- Where will the next pandemic

outbreak occur?

-- Where will transnational criminal activity spread?

-- Where will the next mass migration event occur?

--Where are the populations most susceptible to extremist ideology?

"We would look at a broad range of geospatial information," such as terrain, elevation, roads, buildings, hilltops and rivers, and then consider how people in the area move from one place to another, and over what period of time, Long said.

Local language, ethnicity, education and demographics can suggest whether local populations would be likely to form alliances. And history can reveal previous peaceful or warlike tendencies.

The economy, the climate and access to technology provide additional insight, Long said.

"Our analysis will be greatly enriched by understanding the interrelationship of all [geospatial-intelligence] factors--the Earth's physical features, imagery intelligence and human geography," she said. "The resulting analysis will yield new analytic insights and give the national security community a deeper context to grapple with these difficult questions."

Once they understand the situation on the ground, Long expects her analysts "to communicate this GEOINT analysis visually. Often, the human mind cannot absorb vast amounts of data through the written word alone," she said.

Which brings NGA back to the maps--or their dynamic, interactive digital equivalents.

"NGA thinks spatially and can depict that visually," Long told the symposium. "This is a unique, core competency that we bring to the national security mission."

Human Factors

One thing NGA does not do much of is publicly disclose what it knows about human geography in the places that matter to today's military and political policymakers.

The agency has released few examples of its work influenced by human geography, and none with anywhere near the degree of detail that analysts would typically use. One map--perhaps the only one--cleared for public release is an aerial photo of Baghdad taken in 2009 overlaid with semi-transparent patches of green, orange, red and blue to indicate various concentrations of Shias, Sunnis and Christians.

"Knowing which areas of the city are predominantly Sunni and which are Shia helps coalition forces better understand their environment," says a caption published with the map in NGA's magazine, Pathfinder.

When the same map was published in a commercial imaging magazine, the caption noted that the color patches

represent only "hypothetical areas of different ethnic groups."

In person, NGA analysts are no more forthcoming. When asked how human geography has contributed to handling recent crises, such as the uprising in Egypt this spring that ousted President Hosni Mubarak, two NGA analysts balk.

"I don't know if we should get into that or not," says Craig Rickert, NGA's human geography domain lead. "That's too current, too recent to get into."

His colleague quickly concurs. "I don't want to get into what we were or weren't doing here at NGA; that wouldn't be appropriate," says Carrie Sallaway, a geospatial analyst and member of the agency's vision team that is working to incorporate more human geography into its analysis.

Libya? Syria? No go.

When pressed, Rickert is willing to talk about how human geography contributed to only one major military and diplomatic undertaking--the Dayton Peace Accords, which was hammered out in 1995, eight years before NGA was created.

The Defense Mapping Agency, NGA's predecessor, and the Army Topographic Engineer Center produced 3-D digital maps depicting the physical features as well as the religious and ethnic dividing lines and demographic differences among the people living in war-ravaged areas of the Balkans.

According to the NGA historian's office, mapmakers and analysts used automated cartography, computer-assisted map tailoring and spatial statistical analysis to produce 3-D maps that treaty negotiators could use to "fly" over disputed territories and see the ethnic divides. It persuaded them to draw boundaries that separated the warring Bosnians, Croats and Serbs.

In the years since the Dayton Accords, enthusiasm for incorporating human geography into geospatial intelligence has waxed and waned, Sallaway says. "We've always done some of it, though it wasn't always called 'human geography.' All good analysis looks at other parts of the picture," she says.

The point of human geography is to enhance NGA intelligence by introducing information about human factors.

"I don't think it's a matter of course that it's being used all the time now," Sallaway says. "We're trying to move in that direction, but we don't necessarily have all the support needed to do that."

There are "architecture issues, education issues, data issues and governance issues. Those are the four categories of needs we've identified," she says.

The architecture issues involve computer deficiencies. Human geography education is what NGA analysts need. Data issues range from a lack of data organization to problems with data sharing. And governance involves

setting common standards so that different agencies collecting and analyzing human geography data will use the same terminology so that data can be shared more easily.

NGA has established a human geography working group to begin resolving such problems with "partner agencies," including the State Department, the Census Bureau, the Army and Marine Corps, as well as with U.S. military allies. The goal is to develop standard terminology for "what we should call things," and ultimately, for what human geography data "should look like," Rickert says. "What we're attempting to do is to bring efficiency" to human geography.

So far, though, most NGA analysts have little or no formal training in human geography, Sallaway says. To change that, NGA has added three courses to the National Geospatial-Intelligence College that focus on human geography.

Tapping Academia

The agency does not plan to create a cadre of human geography specialists, but rather to introduce human geography into everyday analysis. "We're talking about regular analysts doing this work," Sallaway says. Each analyst would apply human geography to his or her work "as it makes sense."

A shortage of human geography data poses another problem. "We have a bit, but a relatively small bit, of information regarding human geography" amid the vast trove of intelligence that resides in NGA databases, Rickert says.

There's a smattering of information on the political affiliations, languages and religions of certain locations, he says, but little on economics, significant events, history and "how things have already changed over time. These are all important aspects that we need to add to GEOINT," he adds.

NGA has few of its own human geography experts, so researchers rely heavily on open source intelligence,

according to Sallaway. "You can talk to experts, talk to people who know the area, people who live there, native speakers," she says. On even the most arcane topics, "there are always experts on pretty much everywhere that you can chat with," she adds.

And "you can go online," Rickert says. The Internet is a vast--if unverified--source for human geography data.

NGA's zeal for human geography is part of a broader effort by the U.S. military to apply knowledge from the social sciences to military operations.

Early in the Iraq war, the Army concluded that its soldiers needed greater understanding of local languages, customs, religious practices, ideology and way of life. The Army called it understanding the "human terrain," and by late 2003 it was sending five-member teams of anthropologists, ethnographers and other social scientists to Iraq to study local cultural environments and advise U.S. commanders.

But these human terrain teams proved controversial. The American Anthropological Association warned that participating in the teams could compromise anthropologists' ethics, undermine anthropology as an academic discipline, and endanger both researchers and their subjects.

The harshest critics charged that human terrain teams were an attempt to weaponize anthropology.

A comparable backlash to NGA's use of human geography seems unlikely, Sallaway says. "It depends on whether you're operating on a tactical level or a strategic level. The human terrain teams were in certain towns, operating at a local level." That's not how NGA envisions its analysts using human geography. Mostly they'll be at NGA sites in the United States, and human geography will be one among many factors that shape their analysis, she says.

The question is whether it will work.

Perhaps, says Jerome Dobson, a geography professor at the University of Kansas and president of the American Geographical Society. "All of Director Long's stated questions conceivably can be addressed within the intellectual framework of human geography," he says.

"The challenge, of course, is to pick the right indicators to answer any specific question," he says. In other words, just because two things happen in the same place does not mean that one caused the other.

Picking the right indicators is difficult in the field and far more difficult when working merely with aggregate data at a distance, Dobson says.

Simply gathering human geography data is likely to be hard for intelligence agencies--they're widely distrusted.

"Much of the essential information cannot be collected by the intelligence community or any academic researcher with an intelligence community label on his or her funding," Dobson says.

Another academic observer agrees. "People are nervous about it. No academics want to be portrayed as spies," says an anthropologist at a Midwestern university, who asked not to be identified. He is heading overseas for a summer research project, and says that merely being quoted in a story about NGA could discourage cooperation from local officials.

Real understanding of human geography "comes only from long-term fieldwork in foreign places," Dobson says, and the amount of that work being done by U.S. universities has dramatically shrunk.

"Before 1948, the U. S. government looked to academia for this kind of understanding, and geographers were prominent in the public arena. After 1948, the United States relied almost exclusively on intelligence agencies," he says.

But the spy agencies, some say, haven't done a good job. "American interests at home and abroad have been severely damaged by geographic ignorance about foreign places and peoples," Dobson says. "We made uninformed choices about going to war [in], for example, Vietnam and Iraq."

On the positive side, the military now recognizes its deficit of understanding foreign cultures, and "we see military and intelligence organizations increasingly recognizing human geography as a top priority," he says.

But, he argues, leaving it up to intelligence agencies is not the best way to go.

"What's desperately needed is a joint thrust in which NGA and other intelligence agencies do what they do best--overhead imagery, human intelligence, analyses of massive databases--while the State Department and other civilian agencies fund a massive program to send academic researchers abroad," Dobson says.

Dobson says he and Lee Schwartz, geographer of the United States for the State Department, have been promoting an effort "to recruit and deploy academic geographers and regional experts to places around the world.

"The universal reaction has been enormously positive," he says, but there has been almost no financial support for the idea.

Whether that will change is uncertain, but it's clear that the National Geospatial- Intelligence Agency understands the potential value of human geography.

Says NGA's Long: "If we can use our GEOINT expertise to focus the national security community on an issue before it becomes a crisis, we will have given everyone the opportunity to leverage their assets more effectively, and we will have given the policymaker valuable time to consider a broader range of policy options."%08

William Matthews is a freelance journalist who has been covering government and technology in Washington for two decades.