It’s a scenario straight out of Cold War-era science fiction: A plucky adversary detonates a nuclear weapon dozens of miles above Middle America. The goal isn’t to incinerate St. Louis, but rather to grind daily life from coast to coast to a halt.

A three-pronged electromagnetic pulse hits everything within line of sight of the blast, frying almost every electrical circuit it touches. It’s as if lightning struck every house in every city within a day’s drive at once. The pulse also couples with interstate power lines, both above and below ground, and ripples outward, overloading distant systems. Communications networks go silent, as do Wall Street trading floors, air traffic control towers and intensive care units across the region. Water taps and gas station pumps go dry. Grocery shelves quickly go empty. In the best-case scenario, society laments its over-dependence on electronics and patiently learns how to pick up the pieces in the new dark ages.

Depending on which group of scientists you believe, the threat of a high-altitude electromagnetic pulse, or EMP, attack can range from an overhyped doomsayer fever dream to a grievously overlooked and near-existential threat to the United States – one in which the detonation of a single nuclear device above the Earth’s atmosphere could wipe out 90 percent of the U.S. population (and a sizable chunk of Canada’s, to boot) within a year. The reality, as tends to be the case, is somewhere in between.

The politicization of the issue, along with the highly classified nature of U.S. research into the weapons, makes the EMP issue a difficult one to assess. Nonetheless, with North Korean state media over the past year repeatedly threatening a high-altitude EMP attack against the U.S. – and with the North moving rapidly toward both a high-yield weapon (which is likely needed to pose a major EMP threat to the U.S.) and the ballistic missile capability needed to deliver it – the debate over the severity of the threat is certainly consequential. Given how quickly the crisis on the Korean Peninsula could break in any number of directions at the moment, any potentially relevant factor such as EMP merits scrutiny.

This Deep Dive surveys the state of the EMP debate and attempts to clarify what the points of disagreement tell us about the severity and geopolitical relevance of the threat. It also considers the strategic context in which a country like North Korea may be tempted to launch such an attack. Ultimately, it concludes that EMP is a threat, but the uncertainty surrounding its effectiveness makes it a far less valuable option than a standard nuclear attack. Moreover, however seriously U.S. leaders take the threat, they would respond to an attack as if it were a full-blown nuclear assault. Thus, from a strategic perspective, it would make sense for a “rogue state” like North Korea to conduct a high-altitude EMP attack only under a narrow set of circumstances – albeit ones that could result from the current state of the crisis on the Korean Peninsula.

How It Works

There’s nothing particularly controversial about the basic science of an EMP, and there’s little doubt that an electromagnetic pulse produced by the atmospheric detonation of even a simple low-yield fission device could cause at least some damage to critical infrastructure. After all, the U.S. and the Soviet Union conducted a combined 20 exoatmospheric nuclear explosions from the mid-1950s to the mid-1960s, creating on a small scale some of the problems that are being warned about today. Solar storms and lightning strikes have also caused relatively mild EMPs.

A high-altitude EMP attack would work like this: A nuclear device explodes at high altitude, somewhere between 25 miles (40 kilometers) and 250 miles above the Earth, producing powerful gamma rays that radiate outward. Upon colliding with molecules in the Earth’s atmosphere, the downward-directed gamma rays create a powerful electromagnetic energy field. The EMP doesn’t hurt humans directly, but it makes some electrical devices and attached cables act as antennas, hitting electronic systems with a surge of high-voltage current.

(click to enlarge)

The EMP arrives in three phases – a near-instantaneous, powerful pulse known as E1, a subsequent high-amplitude pulse known as E2, and a slower and lower-amplitude (but still damaging) waveform known as E3. E1 causes most of its damage by inducing voltage in electrical conductors beyond what they can handle. E2 pulses behave similarly to the current produced by a lightning strike, and thus would likely be the least-damaging phase (assuming standard lightning protections haven’t been disabled by E1). E3, which can last from several seconds to several minutes, occurs when the fireball from a large detonation briefly warps the Earth’s magnetic field. Its effects are akin to those of a geomagnetic storm caused by solar flares. It feasts on long electrical conductors, such as power and telecommunications lines, allowing its effects to ripple outward.

The EMP field will extend to anything within line of sight of the explosion, meaning the pulse can, theoretically at least, wreak havoc across thousands of miles, depending on the altitude, the design and the power of the nuclear burst. In other words, a detonation at 60 miles above the Earth could expose an area with a radius of 700 miles on the Earth’s surface to the pulse. Given the interdependence of sensitive networked systems in the United States, such as power and communications systems, the extent of the damage caused by an EMP can cascade outward even farther.

Assessing Its Power

The potential damage that could be done by an EMP is difficult to assess – and thus prone to major disagreement within the scientific and national security communities – in part because it hinges on a vast array of factors. The first, of course, is the power of the EMP, which depends on things like the size of the blast and the altitude at which it is set off. The power of the EMP can also differ from one part of the globe to another, based on distance from the equator or the strength of the magnetic field of the region below. (Generally, the farther the blast is from the equator, and the stronger the magnetic field, the stronger the pulse.)

One area of debate about the EMP threat is whether a simple fission device with a yield of, say, less than 10 kilotons – the sort of device that would most likely be held by a fledgling nuclear state or non-state actor – would really be powerful enough to cause widespread damage. The power of the first phase of the EMP for a low-yield device is believed to be limited by the narrow range of altitudes at which it can be detonated while still causing significant damage. For example, a 1-kiloton device is believed to be strongest if detonated at around 25 miles above the Earth. If detonated much higher, the electromagnetic pulse would dissipate too much. If detonated much lower, deep inside the Earth’s atmosphere, it wouldn’t produce an EMP of consequence at all. (Any high-altitude EMP attack must be conducted at an altitude of 20 miles or above.) At 25 miles, the area of the Earth’s surface within line of sight of the blast, and thus theoretically exposed to the E1 pulse, would have a radius of some 440 miles. Skeptics also argue that E1 from a low-yield device (by some estimates anything with a yield less than 100 kilotons) would also weaken considerably toward the periphery of the exposed region, shrinking the potential area of damage further to a 250-mile radius.

(click to enlarge)

Moreover, low-yield devices produce a disproportionately smaller ionized fireball, which is what’s primarily responsible for the third phase of an EMP. Since E3 is most damaging to long power or communication lines – and the only phase capable of causing significant damage to underground or underwater cables – the current of high voltage wrought by a low-yield EMP is less likely to ripple outward beyond the directly exposed area. (It’s worth noting that a nuclear device could theoretically be designed to produce a “super-EMP.” For example, part of the shielding surrounding the fissile core of a device could be weakened in order to channel gamma rays downward. This can enhance the power of the E1 pulse somewhat, but doing so would not create an E3 pulse.)

By comparison, a megaton-yield bomb can be detonated at as high as 250 miles above the Earth and produce a strong enough E1 pulse to target a vastly larger area. Larger thermonuclear devices (also known as hydrogen bombs) also produce large ionized fireballs needed for the damage-magnifying E3 phase. Of course, larger nukes are much harder for states with rudimentary nuclear programs to both develop and deliver. That said, North Korea is edging toward making the debate about EMPs from a low-yield nuke somewhat moot. Estimates on the yield of its sixth and most recent nuclear test in September range from 70 to 280 kilotons. Notably, though, there’s widespread speculation that the bomb was not a two-stage thermonuclear device, as claimed by North Korea, but rather what’s known as a boosted fission device (considerably easier to develop than a thermonuclear bomb). This matters because thermonuclear devices are needed to create a large E3 pulse.

The vulnerability of various U.S. systems to such an attack depends on countless variables, making credible damage estimates even more elusive. There’s good reason, however, to believe that U.S. infrastructure isn’t exactly resilient to disruption. After all, geomagnetic storms (similar to what would be produced by an E3 pulse) produced by solar flares have already proved disruptive. In 1989, most notably, such an event knocked out power for much of Quebec and shut down trading at the Toronto Stock Exchange.

Part of the threat from a potential high-altitude EMP attack stems from the fact that, while increasing in sophistication, U.S. networks have in some ways also gotten far more fragile. For example, in the mid-20th century, when the only EMP tests to date were carried out, most electronics were designed to operate at a higher voltage than their modern counterparts do. The electronic components they relied on, such as vacuum tubes and induction coils for spark ignition rather than solid-state circuitry, were more likely to be able to handle a surge of voltage induced by an EMP. By comparison, today’s much smaller integrated circuits are believed to be nearly a million times more sensitive to an EMP shockwave.

Power, fuel, communications systems, and food and water distribution have all become much more dependent on electronics – for example, by replacing mechanical fail-safes with electronic ones – further expanding the reach of an EMP. Many critical system nodes have already been shielded to protect against potential sources of interference like an EMP. Still, given the interdependence of these highly networked systems, an EMP wouldn’t necessarily need to fry everything it touches on the ground to have a devastating effect. Rather, taking one critical piece of the system offline could, theoretically at least, have a cascading effect that leads to much broader systemic failure. This was demonstrated with the 2003 blackout in the Northeastern U.S. and parts of Canada, which lasted weeks in some areas and affected some 55 million people. This blackout was triggered by a software bug at a single power station in Ohio. According to a 2008 congressional report, modern transmission system grids have roughly half the standby capacity to tap into during emergencies compared to the 1980s and 1990s.

Potential Damage

Given all the variables involved in assessing the impact of a high-altitude EMP attack, along with the disagreements within the scientific community and the lack of publicly available information from classified U.S. research into the issue, estimates about the potential damage vary widely and tend to be short on empirical evidence.

A 2007 study examining the potential economic fallout of an EMP over the D.C.-Baltimore region indicated a wide range of estimated losses – from as low as $9 billion to as much as $770 billion, depending on how much of the region’s electrical grid and communications systems were shut down. According to the study, the time it could take for economic activity to be fully restored would likewise range from a month to several years.

The most routinely cited estimates come from a pair of assessments put together by the Congressional EMP Commission in 2004 and 2008. The commission had access to classified research and was allowed to conduct some testing of its own in a laboratory environment. Its findings weren’t optimistic. According to the 2008 report on critical infrastructure: “The cascading effects from even one or two relatively small weapons exploded in optimum location in space at present would almost certainly shut down an entire interconnected electrical power system, perhaps affecting as much as 70 percent or possibly more of the United States, all in an instant. … Should significant parts of the electrical power infrastructure be lost for any substantial period of time, the Commission believes that the consequences are likely to be catastrophic, and many people may ultimately die for lack of the basic elements necessary to sustain life in dense urban and suburban communities.”

The following year, the chairman of the EMP Commission told Congress that the damage in areas within the blast radius would be an order of magnitude worse than what Hurricane Katrina inflicted on the Gulf Coast in 2005 – and that a 90 percent fatality rate nationwide within a year due to starvation and systems breakdown was plausible. Since then, public officials, including a former CIA director, have routinely given credence to the 90 percent figure. But experts in the scientific community have dismissed this figure, along with a number of other commission findings, as speculative and/or contingent on factors that are basically impossible to model (the resiliency of modern-day humans without some of the trappings of modern-day life, for example).

Compounding the lack of clarity on the threat is the fact that a high-altitude EMP is almost impossible to test without putting major populations at risk, while the historical record tells us only so much about its potency. For example, in one oft-debated 1962 test, known as Starfish Prime, the U.S. detonated a 1.4-megaton nuclear warhead at an altitude of roughly 240 miles over the Pacific Ocean. To the surprise of the scientists, some 870 miles away in Hawaii, the EMP it produced took hundreds of streetlights offline, damaged some communications equipment and disabled at least three satellites in orbit. This could be interpreted as proof of concept or proof of hype. The device used was far larger than anything North Korea can likely produce anytime soon, and still it did not exactly send Hawaii back to the Kamehameha age. On the flip side, Hawaii was on the outer edge of the exposed area, its communication and electric lines are far shorter than those relied on in the mainland, and interdependence and microcircuitry were both far less of a potential problem in the 1960s. The test tells us only so much.

Perhaps more illuminating was a series of Soviet tests conducted over Kazakhstan at around the same time. A pair of 1.2-kiloton devices (too small for a strong E3 pulse), detonated at 95 miles and 185 miles above the Earth (too high for a strong E1 pulse), caused minimal damage to Kazakh infrastructure. Later, a 300-kiloton device set off at an altitude of 180 miles was believed to be far more damaging, taking out several overhead transmission lines and a key 600-mile underground power line (underscoring the importance of an E3 pulse), while leading indirectly to generator breakdowns and several fires. And, once again, the damage would likely be higher today than in the 1960s. The E1 pulse, however, was purportedly weaker than models would suggest. Available data is sketchy, but it does not appear to have sent Kazakhstan back to the Golden Horde age. The test, again, tells us only so much.

In sum, the damage wrought by low-yield devices would likely be limited both in geographic scope and, since it lacks an E3 component, the type of systems affected. Carrying out such an EMP attack successfully also carries a much smaller margin of error compared to large nukes. An EMP from a large nuclear device is capable of causing extensive damage to U.S. systems across a vast area, but it’s unclear to what degree it actually would – and there’s little publicly available evidence to support the most extreme estimates being warned about. The uncertainties surrounding its effectiveness would give any potential attacker pause before launching one, and likewise will make U.S. leaders hesitant to commit the vast resources that it would take to protect U.S. electronics systems against the possibility of one.

Could North Korea Pull Off a High-Altitude EMP Attack?

Still, even if the EMP threat is prone to exaggeration, it can’t be dismissed altogether – especially given that North Korea is moving toward the sort of high-yield nuclear weapons that have been most successful at generating damaging EMPs in the past. So it’s worth investigating what is perhaps the pivotal question: In what scenarios would it even make sense for a country like North Korea to resort to a high-altitude EMP attack? To put it bluntly, there aren’t many.

The main hurdle to a high-altitude EMP attack is the same argument against a nuclear attack – the threat of retaliatory annihilation. Even if the North were able to trigger an EMP with a nuclear device far more powerful than anything it’s tested to date (delivered via an intercontinental ballistic missile capable of carrying a heavier warhead than any rocket it’s tested to date), and even if the resulting EMP is as strong and effective as feared in the most extreme scenarios, it would not strip the United States of its ability to strike back. Most critical military equipment is hardened sufficiently to protect against an EMP, particularly strategic systems. Some military systems may still be adversely affected by the damage done to connected civilian systems – it’s impossible to test for every conceivable contingency, and certain problems reveal themselves only in a combat environment. But Pyongyang certainly could not assume that enough critical U.S. military systems would go offline to prevent the U.S. from striking back. More problematic, the U.S. nuclear triad – which includes nuclear submarines operating far out of harm’s way – is designed to ensure survivability.

Moreover, a high-altitude EMP attack cannot yet be conducted with any degree of stealth, nor in a way that would give the North any degree of plausible deniability (compared to, say, a state commissioning an attack using some sort of high-yield backpack bomb in Times Square). Until North Korea can make substantial leaps forward in its satellite program, a high-altitude EMP attack has to be conducted with long-range missiles. And to even attempt to paralyze the entire U.S., North Korea would be forced to launch multiple missiles. U.S. sensors would detect the launch and the trajectories would be consistent with a massed missile attack on the United States. American Defense Support Program satellites would detect the origin of the launch, and Pyongyang would have to assume that it would be interpreted as a direct nuclear attack. It would also have to assume that a U.S. retaliatory response would be underway within minutes.

Things go on autopilot from there, leaving no room to change course once it’s determined that the attack is not a conventional nuclear strike but an EMP. Decision-makers would not wait to see if an EMP attack lives up to the hype. (The speed with which this process unfolds undercuts the already dubious theory that a U.S. administration may deem a retaliatory nuclear strike following a high-altitude EMP attack as violating the principle of proportionality and simply stand down.)

So, in reality, the choice facing the North would be no different than if it were deciding whether to conduct a direct nuclear strike on the United States. And if the North gets to the point where thermonuclear war is an acceptable risk, it’s hard to see why it would waste its limited arsenal of nuclear warheads on unproven EMPs rather than on trying to incinerate Los Angeles.

At this point, it would make sense for the North to launch a high-altitude EMP attack in only one scenario: if it were faced with a choice to either use it or lose it.

The main limitation of North Korea’s nuclear program is the unproven state of the re-entry technology of its growing arsenal of long-range missiles. Making sensitive components such as the guidance systems of an ICBM survivable as the missile returns from space to the Earth’s atmosphere is extremely difficult. The North has yet to demonstrate re-entry capability in any of its missile tests, and the way the U.S. talks about the state of the North’s program suggests Washington believes the North isn’t quite there yet. It’s possible that it is and just isn’t showing it, but let’s assume that Pyongyang is still in the exceedingly dangerous window between developing a nuclear weapon and not yet being able to deliver it with any degree of confidence.

If the North finds itself under attack by the United States and determines that it will soon be denied the chance to master re-entry technology, attempting a high-altitude EMP attack may be one of its few remaining cards left to play. This is because a high-altitude EMP attack does not require sophisticated re-entry or guidance technologies. The device is detonated above the atmosphere, after all. In this way, EMP can serve as a sort of bridge between the North’s status as nuclear aspirant and full nuclear power.

Of course, this would still invite massive retaliation from the United States, escalating what may be a concentrated effort to uproot and disarm the current regime to annihilation. Maybe Pyongyang cares about protecting North Korean civilians from such a fate. Maybe it doesn’t. It certainly would like the U.S. to think that the U.S. homeland can’t avoid the consequences of war in any scenario. That’s the logic of credible nuclear deterrence anywhere.

This is why the North has floated the possibility of an EMP attack. It knows the U.S. knows it still needs time to make its traditional nuclear deterrent credible, and it’s looking for ways to stall a U.S. attack until it crosses the Rubicon. An EMP attack may or may not live up to the hype. But the possibility that it could cause serious problems for the U.S., combined with the possibility that the North will empty its chamber when under attack even if doing so invites massive retaliation, gives the U.S. one more factor to consider before moving forward with the military option.