Deterring North Korea

Ground-based Midcourse Defense system faces turning point

Should North Korea’s leaders ever consider launching a nuclear-tipped missile at the United States, there are two considerations that might give them pause: First, there is the certainty that North Korea would be committing national suicide. Second, there is the uncertainty that the missile’s warheads would ever reach Honolulu or Los Angeles.

That’s because the U.S. plans to body-slam any warheads headed this way with darts that would be launched atop boosters sprung from among dozens of silos in Alaska and California.

The task for the $28 billion Ground-based Midcourse Defense system, or GMD, is often compared to hitting a bullet with a bullet, but that’s probably an understatement. First, the interceptor must spot the bullet in the vast expanse of space. Then it must intercept it at a combined closing velocity at least 10 times higher than that of two bullets fired at each other.

No one can credibly promise that the GMD system will work as planned in an emergency, but with North Korea exploding nuclear bombs underground and testing long-range missiles, the U.S. plans to spend hundreds of millions of dollars in 2017 to continue improving a system that was rushed into service in 2004 to meet then-U.S. President George W. Bush’s deadline for deploying a missile defense system.

Intercepting a warhead arcing through space requires fast detection of the missile launch followed by the firing of an interceptor missile armed with an Exoatmospheric Kill Vehicle, plus accurately discriminating the real warhead from what’s likely to be decoys and a cloud of debris left by a warhead or warheads separating from a missile.

Norm Tew, the program director for prime contractor Boeing, notes that the GMD system must link seven kinds of sensors spanning 15 time zones. The sensing needed to detect and knock down warheads comes from the Space-Based Infrared System satellites; Cobra Dane upgraded early warning radars; the Ballistic Missile Early Warning System radars in Greenland and England; Precision Acquisition Vehicle Entry Phased Array Warning System radars in the U.S.; a floating Sea-Based X-band radar; land-based mobile X-band radars; and Aegis ship-based radars. Tew, who has been working in missile defense since 1983, describes GMD as such a vast conglomeration of sensors that it is “the missile system for which you can’t stand in one place and see everything required to commence an engagement.”

As vast an undertaking as it is, it’s clear what GMD is not: an impervious “Star Wars”-esque shield meant to eliminate the specter of total nuclear annihilation. If hundreds of Russian intercontinental ballistic missiles were to streak in over the North Pole, GMD won’t stop them. GMD is strictly aimed at blocking a missile strike from a rogue nation, in particular North Korea, but the system could also defend against Iran, should the international nuclear agreement fail.

Why it’s the focus

Currently, the U.S. has at the ready 30 ground-based interceptors: 26 at Fort Greely, Alaska, and four at Vandenberg Air Force Base, California. By 2017, 44 are scheduled to be primed for launch from their underground silos. In an attack, a volley of them would rise from their silos. Once in space, each booster would release a rocket-propelled metal cylinder. Each cylinder, guided by its own optical and infrared sensors plus targeting data fed from ground- and space-based sensors, would slam into an incoming warhead, the sheer kinetic impact of the collision destroying or disabling the warhead.

The goal is to destroy a warhead in the middle of its approximately 30-minute flight, when the launch vehicle has ascended through the atmosphere and into space and the engines have burned out.

To understand why the Pentagon has made such a focus of going after warheads in space, instead of only when they are conveniently closer to the ground, consider the three stages of an ICBM’s trajectory: boost, midcourse and terminal. All things being equal, experts say the best time to shoot down a ballistic missile would be during the boost phase, when it is ascending slowly on a pillar of fire that makes a lovely beacon for optical and infrared sensors and weapons. In addition, the warheads and decoys are still nestled in the nose cone, so there is only one target. The trade-off is that boost phase is geographically and technically challenging. As a 2004 American Physical Society report pointed out, the boost phase only lasts two minutes for solid-fueled missiles and three minutes for liquid-fueled, which leaves little time for interception. An interceptor, or perhaps someday a laser, would need to be positioned or flown close to the enemy’s launch site.

Then there is the terminal phase, when the warhead is falling through the atmosphere toward its target. Terminal interceptors would have an easier time picking out targets from decoys, because real warheads fall through the atmosphere more slowly and heat up more quickly than heavier warheads sheathed in protective materials. The downside is that the terminal phase might last only 30 seconds, and the warheads could potentially take evasive action or detonate above their targets.

Experts hope that defending the homeland will never come down to a shot in the terminal phase alone, which is why the midcourse is the main focus, at least for now. In the long term, the U.S. Missile Defense Agency wants to put lasers on high-altitude, long-endurance unmanned planes and destroy missiles in the boost phase from standoff ranges. U.S. Navy Vice Adm. James Syring, head of the MDA, said in August 2016 that the agency plans to test lasers aboard MQ-9 Reaper unmanned aircraft. Last year, five prime contractors — Boeing, General Atomics Aeronautical Systems, Lockheed Martin, Northrop Grumman and Raytheon — completed MDA-funded studies to assess the feasibility of an airborne laser demonstrator. In fiscal 2017, MDA plans to award two contracts for preliminary design of a multi-kilowatt laser to be mounted on a high-
altitude manned or unmanned aircraft. The goal is to flight test a prototype in 2020.

Why it’s hard

Development of GMD has been technically challenging, especially the Exoatmospheric Kill Vehicle, or EKV. An MDA fact sheet on GMD lists 17 tests between 1999 and 2014, of which nine were deemed successful: Three of the eight failures were caused by the kill vehicle not separating from the booster; two involved sensor issues on the kill vehicle; two involved failure to launch due to problems with launch software or silo hardware; and one was scrubbed because the target vehicle malfunctioned.

Ted Postol, a professor emeritus of science and international security at the Massachusetts Institute of Technology, argues that the basic science behind GMD is flawed.

Postol believes that while the kill vehicle’s sensors can detect objects in space, they can’t discern warheads from decoys until it is too late to intercept. Assume the warheads and decoys are traveling at around 7 kilometers per second, and the kill vehicle at around 8 kilometers per second, for a combined closing speed of 15 kilometers per second, Postol postulates. If the kill vehicle’s sensors only register indeterminate points until the targets are about 10 kilometers away — his best guess based on likely fields of view and sensor dimensions — then once the true target is discriminated, there would be less than one second to adjust course and strike it.

Postol compares the task to a street-corner shell game: You can see the hustler whirl his three cups over the table, but you can’t be sure which cup has the little ball. Picking out the warhead from among a cloud of decoys and debris can’t be done quickly enough, he says. The only way to find the real warhead would be to have advance knowledge of the characteristics of the warhead, such as its shape, temperature and color.
MDA and the companies that build GMD have a hard time blunting such arguments with specifics, because they fear that disclosing technical details could enable an adversary to spoof or evade a kill vehicle.

Still, Air Force Brig. Gen. Bill Cooley, GMD program director at MDA, expresses confidence about the GMD system. “Objects have different [sensor] signatures,” he tells Aerospace America. “We use all phenomenology to perform discrimination.”

Tew points out that GMD uses a combination of technologies, including infrared and visible-light sensors on the EKV, plus ground- and space-based sensors that feed updated targeting information to the booster and kill vehicle in flight.

“With any one type of technology, you can figure out how to confuse” a kill vehicle, Tew adds. “So the key is you want to use all the types to make it really difficult for anything to get past it.”

Even if the EKV’s sensors work as designed, hitting a fast-moving warhead will require the kill vehicle to maneuver extremely rapidly. Which is where the GMD story gets especially complicated. A persistent problem with GMD has been rough combustion of the EKV’s thrusters. This shakes the kill vehicle’s initial measurement unit, which must determine the kill vehicle’s position relative to the target. In at least one test, the shaking caused the kill vehicle to miss its target.

That problem affected the first generation of the EKV fielded in 2004, called the Capability Enhanced, or CE-1. New interceptors are equipped with the CE-2 models whose inertial measurement units are cocooned against vibrations caused by rough-firing thrusters. Engineers also improved the sensors, electronics and communication components. MDA declines to specify the exact mix of CE-1s and CE-2s in the field. However, the agency says Redesigned Kill Vehicles currently under development will replace all existing CE-1s by fiscal year 2022. The CE-2 was tested Jan. 28, 2016, in what MDA called a non-intercept test. The kill vehicle wasn’t supposed to hit the target but rather get close enough to show that its sensors and thrusters worked. MDA proclaimed the test a success, but in July, the Los Angeles Times reported the kill vehicle had not homed in anywhere near the target. MDA maintains that the test was not meant to be an intercept and that it was successful.

The January launch is not listed on the MDA fact sheet describing test results, and MDA says this is because the sheet lists only intercept tests. The agency provided a list of 11 non-intercept tests between June 1997 and January 2016, all of which were described as “achieving test objectives.”

Early deployment

President George W. Bush in 2002 ordered the Pentagon to put a GMD defense in place by 2004. In his January State of the Union address, Bush had accused North Korea of “arming with missiles and weapons of mass destruction, while starving its citizens,” and he placed its government in what he called an “axis of evil” with Iraq and Iran. For acquisition officials, Bush’s decision meant that GMD had to be developed at the same time as it was being fielded. Bush’s predecessor, Bill Clinton, had started the GMD program by signing the National Missile Defense Act of 1999, but Clinton had deferred a deployment decision to his successor.

In subsequent years, Government Accountability Office reports criticized the Pentagon for deploying equipment before it was fully tested. As a 2012 GAO study noted, while “some concurrency is understandable, committing to product development before requirements are understood and technologies mature or committing to production and fielding before development is complete is a high-risk strategy that often results in performance shortfalls, unexpected cost increases, schedule delays and test problems.”

MDA continues to work toward improving the system. The agency wants $274 million in fiscal 2017 for the Redesigned Kill Vehicle. Another $72 million would go for development of a Multi-Object Kill Vehicle. Just as ICBMs can carry multiple warheads, a single interceptor would carry multiple kill vehicles.
The agency also wants an upgraded interceptor that could be launched as a two- or three-stage booster depending on the range to the target.

As program director Cooley sees it, the GMD program has reached a turning point. The focus has shifted from basic development to making the system reliable and sustainable. A CE-1 built in 2004 is now 12 years old, raising issues of obsolescence and maintaining an industrial base for spare parts. Cooley wants to see a kill vehicle “that can last for decades.” ★

Related Topics


About Michael Peck

Michael writes about defense and space technology. His work has appeared in Foreign Policy magazine, The National Interest and C4ISRNet magazine.

“With any one type of technology, you can figure out how to confuse [a kill vehicle]. So the key is you want to use all the types to make it really difficult for anything to get past it.”

Norm Tew, program director for prime contractor Boeing


Each Exoatmospheric Kill Vehicle, or EKV, is a cylinder with four thrusters encircling
its midsection. The kill vehicle’s three sensors — two infrared and one that detects visible light —
must pick out a nuclear warhead or warheads from amid a cloud of decoys and debris from its booster. Once the sensors pick out the warheads,
its thrusters line it up to intercept the target.

Nighttime view of earth from space showing city lights and part of a satellite solar panel on the left.
North Korea and South Korea as photographed from the International Space Station. In 2004, then-President George W. Bush accused North Korea of “arming with missiles and weapons of mass destruction, while starving its citizens.” Credit: NASA
Engineer in a cleanroom suit inspecting a sophisticated piece of machinery with various technical components and labels.
A Raytheon engineer conducts final inspections during assembly of an Exoatmospheric Kill Vehicle. Credit: Raytheon
An aerospace engine surrounded by parts and housing, emphasizing complex mechanical components.
An Exoatmospheric Kill Vehicle is shown in the shroud of a Ground-based Interceptor. Credit: Raytheon

Deterring North Korea