Every man, woman and child lives under a nuclear sword of Damocles,
hanging by the slenderest of threads, capable of being cut
at any moment by accident or miscalculation or by madness.
-John F. Kennedy
In the last sixty years there have been a number of events, accidents or mistakes that almost triggered Global Thermonuclear War. There are many famous examples of international incidents where a chain of events could have lead to devastating outcomes, or there was a threat by one side of that occurring. Cuba, the Suez Crisis, Yom Kippur and several more of similar notoriety, where disaster was narrowly averted by equal turns of dumb luck, diplomacy and presence of mind.
What are less well-known are the few crises that started at low level; junior officers being alerted by an alarm or an unusual reading, consequences quickly escalating to an international scale. Times when it appeared that a nation was under attack by another; sequences of events which could have triggered a massive retaliation and the following suit of every other suitably-equipped nation. Basically the end of civilization.
In each incident, military officers had data that indicated their country was under nuclear attack from another. They had a few minutes in which to decide whether or not to recommend that their commanders, their commanders and their commander launch a retaliatory strike, knowing enemy missiles may already be on their way. Ultimately the data was shown to be misinterpreted or the result of accidents, nonetheless mass mortality got closer to the surface. Several such incidents occurred and no doubt, as the passage of time declassifies relevant documents, others will come to light. Never, however, underestimate the power of a high-ranking official trying to cover his own ass. Undoubtedly many more incidents occurred than will ever come to light, on both 'sides' of the Cold War line.
This writeup is a rough chronology and accounting of the known 'near misses' thus far. It does not include occasions where negligence caused the risk (for example, lax security on a nuclear missile silo or Russia's current alleged distribution of nuclear launch codes to lower-level officers), rather those instances where confusion and/or misinterpretation of events nearly, or could have taken things past the point of no return.
In any given year there are up to one hundred Missile Display Conferences held by military officials in the U.S, to evaluate the threat to the country of unidentified radar tracks. Two further successive conferences occur, with increasingly high-level officials, if the threat is deemed sufficient: a Threat Assessment Conference, and a Missile Attack Conference. The latter is essentially the instigation of World War III and has never occurred as far as is publicly known. Virtually no events have passed the Missile Display Conference stage, though some of those noted here probably did.
It is important to remember that although several of the incidents recorded here were 'local' - that is, the result of problems on one particular side only - that the reaction to those misinterpreted sightings, if unchecked, would eventually have been observed by the opposing side. That side would then have stepped up their own readiness and possibly launched bombers/whatever in response, which only appears to confirm and compound the original mistake, causing a further response and so on and so forth. Something like Clancy's The Sum of All Fears, if I recall correctly. Misinterpretation of events resulting in a growing cascade of responses like this were the reason for the eventual installation of the 'hot line' between the White House and the Kremlin. It allows the two leaders to speak to each other directly to avoid misunderstanding of their respective actions. How many times it has been used, or if it has been used, is unknown and I presume such information to be classified.
Only one of my sources records this event, though I have found obscure and undetailed references elsewhere. Then-president of IBM Tom Watson, Bob Galvin of Motorola and another unnamed executive were on a visit to NORAD Headquarters. They were shown part of the threat warning system which consisted of five lights which lit up when objects were detected on a graded scale from one light for 'normal' up to five lights for 'incoming attack almost certain'. As they watched, one by one all five lights lit up..
"The numbers continued to move, reaching 4 and triggering the giant defense command into action.
'When the numbers rose to four, Peterson [one of the executives] said, 'Key NORAD generals came running from their offices. They converged on the room - it seemed to take only seconds."
The executives were escorted to another room where they waited with considerable disquiet, for about 20 minutes until they were informed the alarm had been false. I have not been able to find out what caused this alarm, so if anyone could help there I would appreciate it. Apparently Reader's Digest ran an article on this incident. The date given is possibly wrong since NORAD wasn't officially established until 1958, though presumably these executives could have been touring as part of the construction and systems integration process.
On this day the Missile Warning Squadron at Thule AFB, Greenland, detected incoming nuclear warheads. The threat warning was automatically passed to NORAD, which proceeded to prepare retaliatory nuclear strikes. Minutes before the launches were to occur it was discovered that a computer error at MWS had removed two zeroes from the detected target's ranging information, causing the system to see the moon - 250,000 miles away - as only being 2500 miles away and thus potentially an incoming ballistic missile attack. Long range detection systems were in their infancy at the time and quite prone to errors such as these. Several of my sources mention that many errors like this one have occurred - misidentifying such things as the moon, geese and meteor showers as incoming ballistic missiles - that are too numerous to document (they're saying that, but I'm saying it too).
At some point during this day all communications were lost between Strategic Air Command (SAC) HQ and North American Aerospace Defense Command (NORAD). This cut SAC off from the Ballistic Missile Early Warning System (BMEWS) sites. Since the communications were served by triply-redundant and independent telephone lines the assumption was that either all three had failed at once or that NORAD had been affected by 'enemy action', to quote a source.
All SAC bases in the US were put on alert, and bombers were sat at the end of their runways with engines running. A B-52 on airborne alert was eventually used to establish contact with all three BMEWS stations - one in Fylingdales, England, one at Thule AFB in Greenland, the other in Clear, Alaska - to confirm they were still there and operational.
Ridiculously enough all of these redundant communications lines were routed through the same relay station in Colorado. An electrical fault there had caused all three lines to be cut off.
A bear sniffing around a perimeter fence at Duluth Sector Direction Center in Minnesota triggered intruder alarms at all airbases in the area. Unfortunately at Volk Field Wisconsin the alarm was wrongly wired, causing the klaxon ordering out nuclear-armed F-106s to sound. Pilots knew there would be no drills in the current situation so immediately prepared for take off. A staff member had to drive out to the runway to prevent the aircraft from taking off.
A U2 aircraft on an air sampling mission got lost on a mission over the Arctic and penetrated Soviet airspace near the Chukotski Peninsula. Position measurements could only be made with a sextant and visibility was poor, resulting in the pilot drifting too close to the Soviet Union even though pilots had been pre-warned not to go within 100 miles of the border. Soviet MiGs were scrambled in pursuit as the U2 was believed to be a bomber.
The U2 ran out of fuel over Siberia, U.S. F-102s armed with nuclear missiles escorted it on its glide back to safe airspace. The pilots were under orders to prevent the MiGs entering US airspace at all costs and had release authority on their weapons. The U2 was able to land safely without further incident.
On this day an exercise was run at Moorestown, New Jersey, simulating a missile launch from Cuba at the US. At the same time as the simulation was being run a satellite crossed the horizon, causing confusion as to whether it was a missile or not. The facility that should have notified Moorestown of the passing of the satellite had been put on other duties during the crisis.
Radars which could have confirmed or refuted this detection were for some reason not operational at that time, further adding to the uncertainty. For some reason I have not been able to discover, no response was made before the 'missile' made its impact and nothing happened.
A tentative, UN-sponsored ceasefire was in place in the Arab-Israeli war. The day before, fighting had broken out between Egyptian and Israeli troops in the Sinai desert. It was believed that the USSR would try to provide assistance to the Egyptian troops so U.S. armed forces stood up to Defcon 3, the resulting troop and hardware movements intended to be observed by the Soviet Union as a warning against such action. Under such circumstances people get rather twitchy.
So when a technician repairing the alarm at Kinchloe AFB accidentally set off the entire base alarm system, pilots at the Michigan base believed war had begun and rushed to their nuclear bombers. Engines were started in preparation for takeoff but the mishap was realised before takeoff occurred and the aircraft were stood down.
On the morning of this day U.S. Senator Charles Percy was being given a guided tour of the NORAD facility at Cheyenne Mountain, Colorado. While he was there 1000 (yes, one thousand) inbound Soviet ICBMs were detected:
"...at NORAD...all hell broke loose; they were absolutely convinced there were missiles coming at us."
-as Chairman of Senate Committee on Foreign Relations, 96th Congress, 2nd session 1981
This incident triggered the works. The Secretary of Defense was awakened in the small hours with news of the impending attack. Bombers were launched, fighter aircraft were placed on alert and retaliatory missile strikes were readied. The President's emergency airborne command post was even launched - bizarrely, without the President.
Before notifying the President and asking whether he wished to order a counter-attack, Secretary of Defense Brzezinski ordered some final checks. It was then discovered that other radar stations that would have been able to verify NORAD's readings had not in fact detected anything unusual.
Later, it was discovered that a training tape which contained a scenario of a Soviet SLBM (submarine-launched) attack on U.S. bomber bases had been inserted into one of NORAD's computers. This had caused the (operational) computer to 'think' that an attack was really under way, alerting the relevant parties. Subsequently an off-site training facility was established so that training programs would not be run on operational military computer systems.
This incident is probably the most well-known, and occurred at the same installation as the previous one. At 2.26am EDT, Secretary of Defense Brzezinski was alerted that early warning systems indicated two hundred-and-twenty nuclear missiles were inbound to the U.S. A short time later this figure was increased to two thousand, two hundred. It had begun with a detection of just two SLBM launches, ending with an indicated all-out ICBM attack from the Soviet Union. Once again bombers took off, the Pacific Command Airborne Command Post was launched and the Minutemen ICBMs were warmed up.
The alert was rescinded because again, radar readings from across the country did not agree with what was being displayed at NORAD. It took three days to isolate the problem, because it was that long before it occurred again. The fault was traced to a computer chip (which several articles mention cost 46¢, like I care), which was failing randomly. The fault was replacing readouts of incoming missiles that were usually zero with the number two.
On this day a Russian defence satellite tracing a Molniya orbit encountered an unusual planetary alignment which caused it to misidentify natural phenomena as a missile launch from the U.S.
The Molniya orbit is intended to give surveillance satellites an easy time watching for missile launches. It keeps the satellites far away from the earth for extended periods, taking an orbital path such that the area being observed is right on the 'edge' of the world from the satellite's point of view. The exhaust plumes of any missiles launched from that area will show up brightly against the black background of space.
This orbital pattern was (and is) used because it makes it easier to observe extremes of latitude (as compared to geostationary satellites) and makes false readings less likely. The greater amount of atmosphere that light and heat must travel through to reach the satellite's sensors means there's less chance of mistaking glare for missiles.
Nonetheless on this occasion, the position of one particular satellite, together with the earth's alignment with the sun caused the satellite's infrared sensors to identify sunlight reflected from high-altitude clouds as the exhaust trails of several nuclear missiles launched from the U.S. Five of them to be precise.
Colonel Petrov, who was in charge of the observation centre for Russia's Molniya satellites received the warning message shortly after midnight. First one launch was detected, then another, then one after another counting up to five. The amount of 'launches' detected meant the computer system had automatically issued a warning to the Soviet General Staff before the accuracy of the readings could be determined.
After several tense minutes of exchanging messages with subordinates and other officers (not to mention considerable badgering from the General Staff for him to make a decision) he concluded the warning to be false, for several reasons. First, that he had been informed several times of the U.S.'s nuclear doctrine, which at the time was a single, huge nuclear strike. He noted his thoughts at the time, which were something along the lines of: "You don't start a war with five missiles." Second, the satellite system that reported this 'launch' had only been in place for a short time so he was not ready to instantly trust it. Finally, the ground-based radars around the Soviet Union that monitored the horizon for missile launches reported no contacts, even allowing for the several-minutes delay between satellites detecting a launch and the ground radars detecting the same.
Subsequently the Soviet Union established a single geostationary satellite to act as backup for the Molniya satellites - U.S. missile fields could be observed from two very different angles, reducing the chances of a false alarm like this to virtually zero.
If you've not heard of any of the other incidents in this writeup, you've probably heard of this one. It's certainly the only one I personally remember. On this day a NASA rocket was launched from Norway, to conduct a study on the Aurora Borealis. Unfortunately the notification to the Russians of this launch - which was sent several weeks previous - failed to claw up the bureaucratic chain to either the President or the organisation in charge of monitoring for nuclear launches.
Imagine the shock when Russian radars detected an object behaving exactly like a submarine-launched Trident missile. Although the rocket didn't have the same number of stages, it jettisoned stages at more or less the same time as a Trident would during the slice of time for which it was visible to the radar. The angle of ascent and the point of jettisoning stages were remarkably similar.
This warning shot up the chain of command to Boris Yeltsin, who for the first time ever activated his nuclear football, a suitcase that is kept near him at all times with a big red button in it. For fifteen rather tense minutes, discussions took place and more readings were taken to determine whether the US had, in fact, launched a surprise missile attack on Russia.
After about eight minutes of deliberation and discussion with flustered radar operators, it was determined no attack was under way as the rocket's path would not take it inside Russia. Still though, this is illustrative of the types of confusion that can occur. An unidentified Russian diplomat is reported to have stated that many other incidents of this nature have occurred in Russia. I will end the writeup on that thought.
wertperch says re accidental nuclear war: So Peace on Earth, eh?
JoeBaldwin says re accidental nuclear war: This would be funny if it didn't almost result in the untimely death of millions of people...
- Kick, Russ; "Disinformation Book of Lists"; printed word, published by The Disinformation Company Ltd, ISBN 0-9729529-4-4 (this wasn't actually a source, but having read parts I see some of my other sources may quote from it)
- Forden, Geoffrey; "False Alarms on the Nuclear Front"; <http://www.pbs.org/wgbh/nova/missileers/falsealarms.html>
- Rickman, James; "Los Alamos researcher...test monitoring system"; <http://www.eurekalert.org/pub_releases/2003-02/danl-lar021203.php>
- Forden, Geoffrey & Podvig, Pavel & Postol, Theodore A; "The radar and satellite networks...accidental nuclear war"; <http://www.armscontrol.ru/start/publications/spectrum-ews.htm>
also available at <http://www.russianforces.org/podvig/eng/publications/sprn/20000300ieee/index.shtml>
- various authors; "Military/False Alerts of Sudden Nuclear Attacks"; <http://mt.sopris.net/mpc/military/false.alerts.html>
- Philips, Alan F; "20 Mishaps That Might Have Started Accidental Nuclear War"; <http://www.nuclearfiles.org/kinuclearweapons/anwindex.html>
- Hoffman, David; "'I Had A Funny Feeling in My Gut'"; <http://www.washingtonpost.com/wp-srv/inatl/longterm/coldwar/shatter021099b.htm>
- Blair, Bruce G; "Keeping Presidents in the Nuclear Dark"; <http://www.cdi.org/blair/launch-on-warning.cfm>
- various authors; "Stanislav Petrov"; <http://en.wikipedia.org/wiki/Stanislav_Petrov>
- Forden, Geoffrey & Podvig, Pavel & Postol, Theodore A; "Colonel Petrov's good judgment"; <http://www.armscontrol.ru/start/publications/petrov.htm>
- New England Journal of Medicine (author not specified); "Accidental Nuclear War -- A Post-Cold War Assessment"; <http://www.nyu.edu/globalbeat/nuclear/nejm0498.html>
- BBC News (author not specified); "Greater risk of accidental nuclear strike"; http://news.bbc.co.uk/1/hi/world/americas/85881.stm
- Balir, Bruce; "Avoiding Nuclear Anarchy: Containing the Threat of Loose Russian Nuclear Weapons and Fissile Material
- Knelman, F.H; "Y2K and Accidental Nuclear War"; <http://www.peacemagazine.org/archive/v15n3p18.htm>
- Mauro, Ryan; "History of Nuclear Close-Calls"; <http://www.worldthreats.com/general_information/Worlds%20Nuclear%20Close%20Calls.htm>
- Hippel, Frank von; "How to avoid accidental nuclear war"; <http://www.thebulletin.org/issues/1990/j90/j90vonhippel.html>
- Kennedy, Bruce; "Nuclear close calls"; <http://www.cnn.com/SPECIALS/cold.war/episodes/12/spotlight/>
- Yorkshire CND (author unknown); "Fylingdales - False Alarms"; <http://cndyorks.gn.apc.org/fdales/fdalarms.htm>
- PBS Online (author unknown); "Interview - Admiral Stansfield Turner"; <http://www.pbs.org/wgbh/pages/frontline/shows/russia/interviews/turner.html>
- Hoffman, David; "Cold War Doctrines Refuse to Die"l; <http://www.washingtonpost.com/wp-srv/inatl/longterm/coldwar/shatter031598a.htm>
- Tiwari, Jaya & Gray, Cleve J; "U. S. NUCLEAR WEAPONS ACCIDENTS"; via google cache; <http://tinyurl.com/47kzg>
I have long wanted to post a writeup on Christmas Day, though I don't think I wanted it to be so unsuited to the occasion.