Nuclear Mishap. Image: RJHaas. CC BY-SA 3.0, via Wikimedia CommonsMilitary tensions between Washington and Pyongyang continue, and are being sharpened by fiercely provocative statements on both sides. The latter include Trump's implication that the North Korean leadership would not last long, a boast likely to be seen as a direct threat of assassination, and the DPRK's threat to explode a nuclear weapon in the Pacific as a demonstration of its power.

A cluster of recent columns in this series explores some of the issues (see "North Korea: the art of the deal" [3 August 2017], "Trump vs North Korea: a 1945 echo" [10 August 2017], "Trump in a fix: North Korea and Iran" [15 September 2017]. Two central points emerge. The first is the risk that at a time of high tensions, quite untoward events can tip things over into a crisis or even open conflict. The acronym AIM (accidents, incidents and mavericks) summarises the kind of problem that can occur. It must always be kept in mind that military thinking invariably incorporates worst-case planning and even the willingness to “get your retaliation in first”.

The second is that there is now a fundamental conflict of interests between the North Koreans and Trump. For the Pyongyang regime, developing a small nuclear force that can threaten the United States is absolutely essential. Otherwise the leadership believes it has no future. This dates back to George W Bush's declaration in 2002 that North Korea is part of the “axis of evil”. The stance was consolidated by the US-led regime terminations in Afghanistan and Iraq, and the US's involvement in getting rid of Gaddafi in Libya.

The current state of tension worries many people in Europe, but that worry is mixed with feeling that no one would be stupid enough to start a conflict that could so easily lead to nuclear war. This notion stems from a surprisingly persistent impression that nuclear deterrence throughout the cold-war confrontation was robust and stable. On such slnder grounds rest the comforting belief that nuclear weapons kept the peace for more than four decades.

On the edge

A briefing from Oxford Research Group serves as a reminder that key parties to the cold war could actually see a value in a nuclear strategy that included first use of nuclear weapons, and that small nuclear wars could be fought and even won (see Limited Nuclear Wars: Myth and Reality, ORG, 29 August 2017). This attitude persists to the present day. But there is another critique of nuclear stability which deserves revisiting – the largely forgotten experience of crises that so nearly went "hot".

A fearful example of such crises took place on 26 September 1983, and involved a Soviet military officer whose brave action prevented what could have been an accidental nuclear war. Lieutenant-Colonel Stanislav Petrov was the 44-year old duty officer at the command centre near Moscow when satellite elements of the Soviet ballistic-missile early-warning system showed that five United States Minuteman intercontinental ballistic-missiles had been launched from a base in the US midwest and were heading for the Soviet Union.  

Petrov had only seconds to decide whether to report this to his superiors, who could have ordered immediate nuclear retaliation. He did not do so, but instead – acting on instinct – reported a system malfunction. He later said that his gut feeling was influenced by his mistrust of some elements of the early-warning system, and by recollection of a training estimate that any US surprise attack would most likely involve far more than just five missiles.

Petrov was right: ground-based radar confirmed the absence of a US launch. But he was still reprimanded for not delivering a full report. His role was publicised only years later, and by the time he died in May 2017 he had received acclaim and several international awards.

More than a decade later, in January 1995, Russia's early-warning system mistakenly identified a US nuclear-armed submarine-launched ballistic-missile; in reality, the object was a Black Brant XII research-rocket being launched off Norway's northern coast to probe the aurora borealis. This time, the control staff did report to higher authority, and the initial stages of preparing for retaliatory action were begun. Fortunately the mistake was detected before catastrophic damage.

On the American side, one of the most extraordinary episodes came during the Cuba missile crisis in October 1962. The US was then just starting to deploy its new silo-based Minuteman ICBMs, with one site just becoming operational the Malmstrom airforce base in Montana. The launch system was meant to be absolutely secure, so that Malmstrom's launch instructions could only come from the president. But officers at the base calculated that a surprise Soviet nuclear attack on Washington would mean that they couldn't rely on receiving such orders – so they jerry-rigged some of the new missiles to be able to fire them without authority if need be (see Scott D Sagan “More will be worse”, in Sagan & Kenneth N Waltz, The Spread of Nuclear Weapons, 1995). 

There are several other examples, as well as far more cases of accidents involving nuclear weapons. Here are just a few:

* March 1956: a B-47 bomber was lost on an overseas flight, the plane crew and two nuclear weapons were never found

* May 1968: a USS Scorpion attack-submarine reportedly equipped with Subroc nuclear-armed anti-submarine missiles was lost 700 kms southwest of the Azores 

* January 1966: a mid-air collision between a B-52 bomber and a refuelling tanker near Palomares in Spain. Two of the four nuclear weapons were recovered intact, but two broke up and contaminated over 1,400 tonnes of soil and vegetation, which were then removed to a safe-storage site in Texas

* January 1966 (four days later): another B-52 crashed in Greenland, with four H-bombs destroyed, contaminating 1.5 million gallons of ice and snow

* April 1970: a Soviet November-class nuclear-powered attack-submarine – a class of boat known to carry nuclear-tipped torpedoes – sank 300 kms southwest of the coast of Cornwall

* October 1986: a Soviet Yankee-class submarine carrying sixteen nuclear-armed missiles was lost in the Atlantic.

Again, these are selected examples. There are over forty accidents recorded as having affected United States and Soviet nuclear weapons. Many of them are detailed in the  study Too Close for Comfort: Cases of Near Nuclear Use and Options for Policy (Chatham House, 2014). 

At the time of such incidents, a common response of the authorities was to downplay the dangers. In this respect, perhaps the most revealing detail is an almost farcical incident in 1987 when a Minuteman ICBM at Warren airforce base in Wyoming was shown by a computer system to be preparing to launch itself. The authorities later insisted that there was never  risk of an accidental launch, but admitted that officers at the base took the precaution of parking an armoured car on the lid of the silo just in case! (see “Vehicle Parked on Silo After Launch Signal”, Washington Post, 29 October 1987).

The point of retracing this disturbing history today is to reinforce the need to pacify not amplify the Washington-Pyongyang tensions. In addition to all the other risks, that of an AIM-like untoward happening must be recognised. Everything possible must be done to demand that leaders act with wisdom. Especially when that quality seems in short supply at the very time it is most needed