Putting Geography Back in GEOINT

As the geospatial intelligence community expands its capabilities and influence around the globe—and even in emerging virtual worlds such as the metaverse—its leaders would be well-served to incorporate some lessons from a much older, related discipline: geography.

That’s the advice of Lee Schwartz, Ph.D. Geographer of the United States and Director of the Office of the Geographer and Global Issues, U.S. State Department Bureau of Intelligence and Research. Schwartz spoke on Monday at USGIF’s GEOINT 2023 Symposium in St. Louis.

Schwartz considered the challenges of the metaverse from a geographer’s perspective, highlighting aspects of geography that don’t always feature prominently in geospatial or GEOINT analysis. For example, geospatial intelligence is often based on imagery; geographical analysis is more likely to include on-the-ground insight, with substantial weight placed on ground truthing, the practice of confirming at-a-distance insights with in-person measurements. He said he prefers to lean on technologies such as machine learning and artificial intelligence in a supporting role where they provide quality control and assurance, rather than as the primary insight-gathering tools.

“I have to emphasize the value of ground truthing: boots on the ground, working with local partners. Local knowledge is critical,” Schwartz said.

It can also bring complicating factors to the surface. For example, he highlighted an incident in Africa where repatriated refugees returned to their home country after 30 years. The people brought thousands of cattle home with them, only to discover that their traditional grazing lands were being used by other farmers, setting off complex debates over land tenure.

To Schwartz, it illustrates the sort of complexity that a simple border survey could miss, but that good geographers should strive to identify and incorporate. That’s especially true in his work, which includes helping resolve international conflicts over border disputes, as well as facilitating humanitarian crisis response.

The grazing-land dispute could be considered irrelevant to geography by someone with a narrow definition of the field, but Schwartz says that misses the point.

“The real challenge, whether it is in the metaverse or on the back of a napkin, is to be accurate, verifiable, shareable, understandable, useful, usable, and used,” Schwartz said.

Schwartz also cautioned against dismissing ground truthing as a practice relevant to geography but not the tactical domain of geospatial intelligence—especially in military operations.

“A lack of knowledge of a lot of local traditions and customs and other things are that are [more] critical for the warfighter as they are for the academic or the mapmaker,” said Schwartz.

Schwartz’s point about the value of incorporating ground-level insights was part of a larger message for the GEOINT audience: that as the geospatial industry grows, its leaders should take to keep the field rooted in the long-established practices and disciplines of geography, along with the conceptual frameworks that could inform the use of rapidly developing technological capabilities. To that end, he argues that geospatial intelligence is tradecraft within the broader discipline of geography, and the GEOINT field will truly flourish only when its sensors, satellites, and algorithms are rooted within the core principles of cartography and geography.

“We can’t forget the tools and fundamentals of my discipline, geography, even as we move into the metaverse,” Schwartz said. “We have a discipline, we have a canon of thought, we have tried and true methodologies that have been in existence for nearly 200 years.”