Spy satellites enter new dimension

The real revolution in the use of spy satellites is taking place not in the skies, but on the ground. NBC’s Robert Windrem reports on the technology of imagery.

SHARE THIS —

For more than 35 years, spy satellites have roamed the skies 100 miles above the Earth. They’ve become increasingly powerful over the years — but intelligence experts say the real revolution in satellite imagery today is taking place on the ground, not in the skies.

TRAVELING AT MACH 25, they pass over every spot on the face of the Earth twice a day, grabbing digital snapshots of places that the CIA — and the policymakers and military officers it serves — want to see. From missile fields in China or Russia to environmental disaster areas in the Sahel, the spy satellites provide a steady stream of black-and-white images.

Much has been written about these billion-dollar “birds,” five of which are overhead on any given day. Even though their design, code names and capability are classified beyond top secret, there have been enough leaks over the years to sketch out some basic information about them.

Three are “visible light” satellites, the most recent of which resemble the Hubble Space Telescope and were built by the same contractor at the same Lockheed Martin facility in Sunnyvale, Calif. They are known in the spy trade as “Keyhole-class” satellites. And they have a resolution of 5 to 6 inches, meaning they can distinguish an object that small, but no smaller, on the ground. Two other satellites are radar-imaging, built by Lockheed Martin in Watertown, Colo. Their resolution is about 3 feet.

While satellites cannot read license plates, they can tell if a car has one. While they cannot tell a mullah by the length of his beard, they can help analysts figure out how many people are chanting along with him at a street demonstration. And while they cannot hover over an area and provide real-time images, other “assets” such as unmanned aerial vehicles, also known as drones, can do that.

REVOLUTION ON THE GROUND

But the news media have concentrated too often on the space-based element of the reconnaissance systems, controlled from the National Reconnaissance Office headquarters in Virginia. U.S. intelligence officials and private analysts tell NBC News that the real revolution has to do with the manipulation of digital data derived from satellites.

The National Imagery and Mapping Agency, which was set up to centralize that research, development and analysis, has been quietly improving what U.S. officials can “see” through their satellite eyes. At its Virginia headquarters and its main analysis “factory,” the National Photographic Interpretation Center in southeast Washington, NIMA has created a new era in spying.

“It has been the revolution of the ‘soft copy,’ ” says John Pike, space policy expert at the Federation of American Scientists. “By ‘soft copy’ I mean the digital copy.”

Over the past decade, digital data has permitted U.S. intelligence to combine visible light imagery with other imagery to make a two-dimensional image multidimensional. Perhaps more importantly, that digital imagery can be transmitted to users around the world.

For a long time, U.S. satellites have been taking pictures in stereo-optic pairs, meaning side-by-side images taken at slightly different angles. That permits intelligence analysts to get a 3-D view, like you would through a child’s stereopticon.

AN AID TO DIPLOMACY

The value is not just in war fighting, but also in “mission rehearsals” of military and intelligence operations. Even diplomacy has benefited.

The new capability the public is most familiar with — because of the Bosnian peace accords — is the development of 3-D “virtual reality” animations of land features.

During the peace talks, representatives of the warring parties were able to view “Powerscreen” animations of the Serbian countryside derived from unclassified satellite and aerial reconnaissance imagery. The clear message was that combat pilots would be prepared to navigate the real geography unless Belgrade was willing to make a deal.

But there are other uses — some of which are derived from classified imagery — that are even more extraordinary.

CREATING ‘ENVISIONS’

Using initially a Cray supercomputer and now smaller computers, NIMA analysts create 3-D animations — called “envisions” — for policymakers so they can understand problems faced by peacekeepers or soldiers before they make decisions on deployment.

Similar animations were shown to pilots at preflight briefings in 1995 to help them prepare for bombing runs over Bosnia. Multiple route plans have been animated so the pilots know the advantages and disadvantages of each.

And the simulations can be just as useful in the battles against terrorism and narcotics. The CIA has pulled together street-by-street urban landscapes that are used to prepare intelligence officers and agents for missions to “denied areas” like South Beirut before they arrive on scene. With some newly acquired technology, those officers and agents can use a joystick to take a virtual “stroll” through such an area long before they arrive.

Similarly, U.N. inspectors were given an advance look at Iraqi nuclear facilities before their arrival at the actual sites. In fact, U.S intelligence sources say NIMA has a catalog of virtually every critical Iraqi superweapons facility.

Drug Enforcement Agency agents have uncovered the vulnerabilities of well-guarded secure locations, like the hideouts of Cali drug lords, prior to an assault. DEA and Colombian police carried hard copies of the images when they went after the drug lords in 1995 so they could determine routes of escape as well as parapets and other high ground they would have to watch during an assault.

ARCHIVING IMAGES

Pike says that NIMA now maintains an imaging archive that can be accessed via the closest server.

“Until the past few years, the imagery, even though the downlink was digital, had to be converted to film — because physically, the intelligence community didn’t have the bandwidth to move it,” says Pike. “During Desert Storm, an airplane had to fly the pictures to Saudi.”

Now, he notes, there is enough bandwidth to sent these multi-gigabyte images to wherever they are needed.

These 3-D capabilities can even help intelligence analysts determine what a terrorist or drug lord’s intentions might be. For example, if analysts know that a suspected terrorist has rented an eighth-floor apartment in a particular building, they can order a 3-D re-creation of that neighborhood. By “flying” through the neighborhood 80 feet above the ground and freezing the view in front of the suspect’s apartment, the analysts see what the suspect sees — and perhaps gets a good view of what’s being targeted.

BLENDING SOURCES

And the analysts are not limited to satellite imagery. They can add information gathered from other sources to create a more complete 3-D image.

Here are three examples: a CIA agent covertly takes photographs of a “denied area,” like south Beirut. Those photos can be added to the 3-D animation of the neighborhood, created mainly from satellite imagery, yielding a more realistic look for the CIA officer or commando who will follow him.

In the second example, a CIA agent obtains technical data, but not a photograph, on China’s new F-10. When the Chinese roll out a mock-up of the F-10 and a Keyhole satellite snaps its picture, the analyst can take a look at the imagery and then add in the technical data to create a better, fuller and more compelling view for policymakers and military intelligence officers.

Finally, an analyst can combine unclassified, low-resolution, multispectral imagery of North Korea’s nuclear reactor — the kind that shows heat and ultraviolet emissions — with classified high-resolution Keyhole imagery of the same reactor. The commercially available low-resolution imagery can sense a rise in heat — and thus the operational tempo at the reactor — while the high-resolution imagery can watch for the movement of fuel rods or other equipment at the reactor.

Intelligence historian Jeffrey T. Richelson says the U.S. is relying increasingly on unclassified as well as classified data, combining them for a more complete picture of what is on the ground.

“We have become very good at fusing imagery from visible light and radar imaging satellites with imagery from multispectral satellites, which are unclassified,” says Richelson.

Richelson notes that each has its strengths. For example, the classified radar-imaging satellites — initially code-named “Lacrosse” — can see through clouds and at night and to some extent can even see underground. The recent discovery of the ruins of ancient Arabian cities provides the best unclassified example of radar-imaging capability.

Radar images can also be digitally rearranged to create the perspective of seeing the target from all sides, an immense value in the analysis of foreign weapons systems and military installations.

Satellites that have infrared cameras, like the unclassified Landsat, can better detect targets that are camouflaged.

“So, multispectral or hyperspectral imagery can be combined with visible light imagery to detect things that each alone can’t detect,” Richelson said.

MODELING OBJECTS

Another, newer capability of imagery analysis involves “modeling” — creating 3-D computerized models of buildings, ships, planes and other objects, then combining them to obtain further information. One recent example involved manipulating an image of a North Korean freighter to obtain the ship’s internal dimensions, cargo-loading capabilities and maximum load. Then the analyst modeled an image of a North Korean Scud missile. By adding details on the ship’s history and North Korean sales of missiles to other nations, the analyst produced a 3-D model that could help determine how many missiles were loaded on a freighter headed for Iran.

The CIA can now use artificial intelligence along with modeling to match a known building to an unknown location. The analyst and computer scientist can take a covertly obtained blueprint, create a digital model of the facility, then “ask” a computer to scan the available imagery and find the completed facility.

The key to understanding all these advances is knowing the real value of digital imagery: the capacity to manipulate, model and combine. There is much that can be exploited as computers become more powerful — and as scientists and analysts become cleverer.

Robert Windrem is an investigative producer for NBC News.