Joshua PollackGuest Post: Consequences of Testing an H-Bomb in the Pacific

Friend and CNS Scientist-in-Residence Ferenc Dalnoki-Veress has written a detailed primer on what might happen if North Korea detonates a hydrogen bomb over the waters of the Pacific Ocean. We’re pleased to bring it to you here.

Effects of a Potential Thermonuclear Test by North Korea in the Pacific

Ferenc Dalnoki-Veress

For more than fifty years, since the entry into force of the Partial Nuclear Test Ban Treaty (PTBT), almost all nuclear-weapon possessing states have observed a moratorium on nuclear tests conducted above ground. France stopped after 1974, and China after 1980, ending the era of atmospheric nuclear tests. This choice is principally meant to prevent the release of radioactive gases and particulates from bomb debris (known as fallout) into the environment, where it could contaminate large areas after it settles to ground level. As observed by the late Herbert York, the first director of Lawrence Livermore National Laboratory, the PTBT also legitimized nuclear testing and had the consequence of making the “the continuation of uninhibited weapons development politically respectable.” (Quoted in Fehner, Terrence R., and Francis George Gosling. Battlefield of the Cold War: The Nevada Test Site. US Department of Energy, 2006.)

Unhappily, the foreign minister of North Korea, Ri Yong Ho, stated recently in New York that North Korea may test a hydrogen bomb over the Pacific Ocean rather than underground, as they have done previously, ending the global moratorium that has existed since before many of us were born. He did not state explicitly that any such test would involve a warhead mounted on a missile, but this would be a reasonable assumption. If the warhead is a thermonuclear bomb, it may also have an additional jacket of uranium to enhance the yield while at the same time producing more radioactive fallout. What might it look like? What might be the results?

The least provocative version of such a test might be a high-altitude airburst detonation in the region near where the last two North Korean missiles that flew over Japan landed. The geographical location of the detonation is not critical so far out into the ocean, so accuracy of the missile that deliver the warhead is also not critical. (The North Korean missile launched over Japan on August 29 may have landed hundreds of km from where it was expected, based on an analysis conducted by Dr. Marco Langbroek.)

As many have observed, this would be akin to the “Frigate Bird” test of 1962 where the United States Navy tested a 600 kt nuclear weapon coupled to a Polaris SLBM at an altitude of 11,000 feet (3.4 km). The Navy wanted to quell Air Force critics who were not convinced by the “effectiveness of the system under ‘real’ operational conditions.” (See: Spinardi, Graham. From Polaris to Trident: The Development of US Fleet Ballistic Missile Technology. Cambridge University Press, 1994, p. 62.) The North Korean leadership undoubtedly feels pushed into a corner to respond to President Trump’s verbal attacks, but there may also be technical reasons to do a “Juche Bird” test to demonstrate the reliability of the coupled system of missile and bomb. Since a test is also a public demonstration, it may be that North Korea may issue a NOTAM (Notice to Airmen) and/or NTM (Notice to Mariners) over a large region to prevent contamination of nearby ships and air traffic, while at the same time drawing in aerial sampling aircraft to collect information on the nuclear device design. On the other hand, they have only done this for satellite launches, not for missile launches, and to give advance warning might make the missiles vulnerable to a sea-based intercept from U.S. Navy Aegis vessels. So we may not know when they test until they test.

Testing at high altitude

A high-altitude airburst would allow the prompt gamma rays and neutrons emitted by the bomb to be absorbed by the air rather than irradiating ocean water. If the test is conducted in or near water, it would risk the production of sodium-24 through neutron capture on sodium in the salt.  Sodium-24 has a 15-hour half-life and emits fast electrons (beta particles), which may penetrate the skin and cause lymphatic-system cancer. Armed forces personnel were unfortunately exposed to this isotope after the 1946 “Crossroads Baker” test, because they were given permission to swim in the lagoon near to where the test took place without any precautions or sense of how dangerous the water may be. (Robbins, Anthony, Arjun Makhijani, and Katherine Yih. Radioactive Heaven and Earth. London: Zed (1991), p. 7.)

Still, even if the test is sufficiently high to avoid this sort of interaction with water, the intense neutron flux from the rapidly rising fireball would interact with nitrogen in the air and produce the long-lived isotope carbon-14, adding to the inventory already in the air after decades of atmospheric tests. Carbon-14 will become carbon dioxide gas, will be taken up by plants, and will enter the food chain far from where the explosion took place.  It would be difficult to make any causal connection between any single nuclear test and carbon-14 ingestion. However, any new atmospheric test will contribute to the current atmospheric load of carbon-14, whose long lifetime will make it dominant among the residual effects of atmospheric nuclear testing for thousands of years to come.

Fortunately, an airburst test would also be likely to occur low enough so that in the line-of-sight to the Earth’s surface, any electromagnetic disturbance (EMP) effect would occur over a small area and would not affect local air or ship traffic appreciably.

Fallout: early vs. delayed 

Two types of nuclear fallout are of concern: early and delayed fallout. Early fallout consists of particulates that are large and heavy enough that they fall to the surface within 24 hours. This occurs with low-altitude “surface bursts,” which suck material from the ground or the sea up into the nuclear fireball, generating copious amounts of debris that returns to the surface as fallout, as well as low-altitude “airbursts,” where the particulates come from the debris of the bomb itself. Residents of Rongelap Atoll, downwind from the site of the 1954 “Castle Bravo” test, mistook the deadly fallout hours after the test for snow. All 23 passengers aboard a Japanese fishing vessel downwind of the test also got severe radiation poisoning; one died. This early fallout is the most dangerous because most of the radioactive isotopes involved decay quickly. Just 100 hours after a nuclear test, the radioactive dose from fallout is 1/100th of what it was after the first hour. (Craig, Paul P. The Nuclear Arms Race: Technology and Society. McGraw-Hill College, 1990, p. 312.)

Fortunately, a high-altitude airburst would not produce appreciable early fallout. Rather, the fallout expected from a high-altitude test would be delayed fallout only. These are radioactive particulates less than several microns in diameter. Depending on the altitude of detonation and the yield, this fine radioactive dust is injected into the troposphere or the stratosphere. If injected into the troposphere, the particulates will tend to descend to the ground over a period of months, washed out by snow or rain. Particulates that enter the stratosphere, on the other hand, are essentially above the weather. They travel thousands of miles, and can stay airborne for many months or even years until they descend into the troposphere.

The particulates in fallout can be composed of as many as 300 radioactive species known as isotopes. These have varying adverse effects on living things. Depending on the dose, the exposure could lead to severe radiation sickness leading to death, or could cause mutations to DNA that could lead to cancer later in life. Exposure to radiation in utero can result in death of the fetus, microcephaly and/or mental retardation, depending on the gestational age of the fetus at exposure. The effects of fallout do not distribute themselves evenly around the globe, but may be concentrated unpredictably. One such “hot spot” was found in 1953 in Albany, New York, thousands of miles away from the Nevada Test Site, when an intense thunderstorm washed out the particulates as the radioactive cloud passed by. It is unclear how many other places may have been similarly affected by nuclear testing.

The isotopes of greatest concern are iodine-131, strontium-90, cesium-137 and zirconium-95, which together contribute the majority of the radioactive dose. Iodine-131 has an eight-day half-life, and can be inhaled or settle on land where cows graze, contaminating their milk. Once ingested, I-131 concentrates in the thyroid and may cause thyroid cancer. Cesium-137 and strontium-90 have half-lives of approximately 30 years and remain present for about a century, potentially irradiating people through ingestion or external exposure. Strontium-90 is known as a bone-seeker, displacing calcium in bones, which makes it an especially dangerous source of internal contamination. Estimates of the additional deaths due to past atmospheric nuclear testing runs into the hundreds of thousands. (Robbins et al., Radioactive Heaven and Earth.)

What you don’t know can hurt you

It is unlikely that the adverse health effects of any single nuclear test conducted over the Pacific Ocean can be measured, but that does not mean that these effects do not exist and are not important. We are reminded of the late Harvard social scientist Daniel Yankelovich, who identified four fallacies that policy makers make when using quantitative estimates. His third fallacy is particularly relevant here: to “presume that what can’t be measured easily isn’t important” leads to underestimation of the gravity of the situation. An atmospheric nuclear test would be a blow to human welfare around the planet, as well as a departure from a norm that has been observed by all nuclear-weapon possessors for 37 years.


  1. jobrazda (History)

    Well done Ferenc!

  2. robgoldston (History)

    If you take the estimate of 100’s of thousands for the total cancer deaths from all atmospheric testing, presumably based on the linear-no-threshold hypothesis, what is your estimate for a single 200 it test?

    • Ferenc (History)

      Hi Rob,

      Of course this is a very difficult estimate to make which is why I did not want to include it. However, if you assume that the number of fatalities correlates with the yield then if you assume total yield from 1945-1980 was 545 MT and assume 430,000 fatalities (assumption is integration time to year 2000). Then the number of deaths per MT is 788. So if a 250 kt would be conducted the number of deaths is about 200 globally, statistically. We would never be able to tell who died because of the high cancer rate around the world. But as I emphasize it is still 200 unnecessary deaths. We need to de-escalate so it does not happen.

      Again, this is a back-of-the-envelope and obvious can be refined. For details see the document I reference “radioactive heaven and earth” pdf on IEER’s website. Look at page 34-40.

      Kind Regards,

  3. robgoldston (History)

    Thanks, Ferenc, that’s just what I was looking for. When I teach this stuff I am clear that the LNT hypothesis is just that, an hypothesis, but I give the LNT projection along with the caveats.

    It is interesting that strontium was the main fear from atmospheric testing, and cesium is the main concern from reactor accidents. I think this, at least in part, is because cesium is much more efficiently volatized – but that is not an issue for an atmospheric test… everything is volatized.

    It is also interesting that genetic mutation was the biggest concern about atmospheric testing (c.f. Godzilla), but studies of progeny of Hiroshima and Nagasaki survivors have not provided statistically significant data on this, as I understand. But I haven’t read the primary sources on this.

    Just to be 100% clear — I think it would be very bad if NK were to fly a “Juche Bird.” We should be at the negotiating table finding out how we can get a freeze on nuclear and missile tests, so we can then undertake talks for further stabilization. Like the JPOA led to the JCPOA.


  4. J_kies (History)

    Given imperfect missile reliability and realistic expectations. Juchebird is effectively a first strike in a nuclear conflict. Frigate bird did not overfly foreign populations on its way from the Submarine launch in the Pacific to its point of detonation also high over the Pacific. Frigate bird also had safety systems and a safe/arm that precluded function in the event of non-nominal flight.

    An armed nuclear explosive fired from the DPRK to the Pacific holds non-trivial likelihoods of failure to include 1) failure on the pad, 2) failure early in boost leading to ground ranges of less than 200km (and unconstrained dispersal on where the RV could land). 3) Early propulsion failure leading to dropping the RV on Japan near the overflight track. 4) Catastrophic guidance / control failures leading to the missile flying west or north or south putting the PRC, Russia, and ROC, and the Philippines at risk.

    As no foreign knowledge or confidence exists of DPRK flight safety / flight termination systems or weapon systems safe, arm or firing, we must presume any flight by a Juchebird represents a potential nuclear strike on any nation within the flight range.

    That’s the definition of the first shot of a nuclear war.

    • Andrew Locke (History)

      I couldn’t agree more with your logic. Add to all of this the uncertainty of whether it is a test or an actual strike against the US or an ally, and you have what I feel is a very real possibility that the US may choose to destroy any such missile being prepared for launch if it is detected in a timely manner. Kim Jong Un and Donald Trump have threatened each other so much by this point that they may have left themselves with nowhere else to go but after each other’s countries. Self-preservation is a very strong instinct.

    • Simon Gunson (History)

      Nonsense, a missile flight terminated by a nuclear test is not a first strike.

      You are employing distorted use of the English language to lay a false perception. A fist strike is an actual attack on the United States without warning.

      A missile fired into the Pacific with an atmospheric explosion is a timely warning not to mess with North Korea. It is an indication that North Korea means business and can do the deed. Americans need to understand that a nuclear war with North Korea very likely will result in several US cities being wiped off the map with tens of millions killed.

      Why would you argue for entering into such a war in a state of utter ignorance?

  5. Ferenc (History)

    By your (J_kies) definition, a repeat of the launch of Sept 15th is also an act of war. But we could not have been sure before Sept 15th either that the missile did not have a warhead. US didn’t interpret it then as an act of war. What has changed? The statement by FM Ri?

    I think if a missile is launched the US, ROK, Japan should not interpret it necessarily as an act of war precisely because of self-preservation. I think a Juche Bird, is a possible next step along the escalation ladder. And should not be interpreted as an act of war If and only if, the trajectory appears to trace (in the vicinity of) the previous trajectory (not at Guam and not at Tokyo). The US has “terminated” ICBM’s in flight before ( so why should we assume that NK will not do that if the failure modes you describe happen? Even if there is an unintentional failure, US has to remain calm and not miss-interpret as an act of war. This is exactly the kind of miscalculation/incident that scares me. I think it will be clear and we will recognize when the DPRK runs out of steps along the escalation ladder and we should keep cool heads until then.

  6. Anon2 (History)

    “The US has “terminated” ICBM’s in flight before ( so why should we assume that NK will not do that if the failure modes you describe happen? ”

    Because you are dealing with a third world country that may or may not have perfected the engineering in the command destruct package (radios and all); and to prevent an semi-accidental arming of the warhead after an in flight failure.

    The U.S. has never terminated a test ICBM with a nuclear warhead in flight. This overflight is beyond negligent — it is so very dangerous if the “footprint” of the destructed missile parts flies over a populated area.

    It is our hope that the Goldstone scenario plays out: negotiated freeze with eventual denuclearization.

  7. Ferenc (History)

    Thanks Anon2, J-kies and Andrew Locke for your feedback. I think we all agree that we are in a very dangerous territory here.

    I have to ask. Did everything change after FM Ri made his comment? Should we now assume that every missile test could be nuclear tipped and thus preemptively destroy every missile that shows preparations for being launched? Lets analyze this for the hypothetical situation that US detects plans for missile launch which happens time and time again.

    Should the US 1) shoot it down preemptively which is skipping a step on the escalation ladder and would undoubtedly trigger a response from NK and I would argue with high probability lead to war with thousands of casualties. Or 2) ‘grit ones teeth’ and allow the next missile to launch whether it is nuclear tipped or not because it will be impossible to tell (probably next week) at launch? In this case NK moved a step on the escalation ladder. Both options suck and both parties know that, and argues for less likelihood that it will happen! Let’s just hope as you say this scenario will not play out.

Pin It on Pinterest