home | about | documents | news | publications | FOIA | research | internships | search | donate | mailing list
Animated graphic depicting the devastating effects of a 300 kiloton nuclear airburst over the Pentagon as described in an article by Lynn Eden in The Bulletin of the Atomic Scientists. (Graphic animation produced by Rock Creek Creative)

"It Is Certain There Will be Many Firestorms" (1)

New Evidence on the Origins of Overkill

National Security Archive Electronic Briefing Book No. 108

Edited by William Burr

January 14, 2004

Read the press release

Scenarios of nuclear attacks on the United States, whether by terrorist or state adversaries, have assumed that Washington, D.C. would be a major target. A startling article published as the cover story of the January-February 2004 issue of the Bulletin of the Atomic Scientists describes the enormously destructive effects to the Washington, D.C. area if, on a clear day, an adversary exploded, 1500 feet above the Pentagon, a nuclear weapon with an explosive force of 300 kilotons (20 times the explosive force of the Hiroshima weapon). (Note 2) This horrifying scenario, prepared by Stanford University nuclear historian Lynn Eden, shows the weapon's blast effects crushing the Pentagon and knocking down nearby buildings. But the blast effects would be confined to a considerably smaller area than the much larger area set ablaze by the fantastically high levels of thermal energy released by the bomb. Paper, wood, office furnishings, and other common combustibles would be quickly ignited. "Within tens of minutes, the entire area, approximately 40 to 65 square miles--everything within 3.5 or 6.4 miles of the Pentagon--would be engulfed in a mass fire" that would "extinguish all life and destroy almost everything else." The creation of a "hurricane of fire," Eden argues, is a predictable effect of a high-yield nuclear weapon.

Lynn Eden's article draws on a chapter from her extraordinary study, Whole World on Fire: Organizations, Knowledge, and Nuclear Weapons Devastation, published by Cornell University Press in January 2004. The article and the book examine why, even though nuclear weapons reliably produce fire effects, U.S. nuclear war planners never systematically took firestorms into account when they targeted adversaries. According to Eden, a "very conservative estimate" would be that firestorms, also called mass fires, "double the damage" caused by modern nuclear weapons. (Note 3) Yet instead of treating mass fire as a standard effect of high-yield weapons, military planners focused their attention only on blast effects which, Eden argues, led them to underestimate the destructiveness of nuclear weapons. (Note 4) That miscalculation meant that throughout the Cold War and since, the U.S. nuclear arsenal was far more devastating than most civilian and military officials understood. Underestimating the capabilities of nuclear weapons meant that larger numbers of civilian casualties ("collateral damage") became more likely, even if most weapons were aimed at military targets only. This also meant that the U.S. produced far more nuclear weapons than were necessary to achieve the military and political objectives sought by the White House and the Pentagon, thus raising their already formidable costs. (Note 5) In her stimulating study, Eden asks: why did Air Force nuclear planners develop plans and methodologies that predicted blast damage only and systematically ignored the mass fires that were no less predictable?

The blast damage that has received the near-exclusive attention of nuclear planners is one of the 'prompt" or immediate effects of a nuclear detonation. The explosion creates a powerful blast wave that places objects within its path under severe, sharp increases in atmospheric pressure. The increase in air pressure that the blast wave causes is known as "overpressure," the air pressure over and above normal sea-level air pressure (14.7 pounds per square inch [psi]). Attending the blast wave are powerful transient winds that exert "drag pressure" or "dynamic pressure" on structures. Eden's scenario of an attack on the Pentagon scenario demonstrates the impact of overpressure and dynamic pressure. The overpressure at Pentagon City, less than a mile from the detonation, would be 36 psi, approximately the same as that Hiroshima ground zero on August 6, 1945. "The compressed air and winds associated with the shock could cave in buildings and might even topple large office buildings." (Note 6)

Chronological development of an air burst; 3 seconds after 20-kiloton detonation; 11 seconds after 1-megaton detonation. Note the tremendous wind velocity and the high levels of thermal radiation measured in "Cal/Sq CM" (calories per square centimeter), a measurement of energy deposited on surfaces. (Note 7)

As Eden argues, Air Force nuclear planners developed an "organizational frame"--the approach to defining and solving problems that organizations use--that continuously centered their thinking and planning on blast effects. The development of the "blast damage frame," she contends, was a "self-reinforcing organizational processes," much of it "sealed off from the public through secrecy and opacity." It had its origins in World War II bombing campaigns and was reinforced by studies of the damage to Hiroshima and Nagasaki. By the early 1950s, "Physical Vulnerability" analysts working for Air Force intelligence had developed a Vulnerability Number (VN) system that helped predict blast damage with considerable accuracy. (Note 8) The nuclear test series of the 1950s helped nuclear planners refine their thinking about the VN system, but they were not interested in examining or predicting damage from firestorms. Although PV officials recognized that nuclear explosions could cause mass fires, they believed that they involved too many imponderables to make them a predictable phenomenon. Nuclear testers even took pains to ensure that structures did not catch fire so they could better measure the effects of blast. Eden details how the famously photographed House No. 1, destroyed by blast in nuclear weapon test "Shot Annie" on 17 March 1953 was made highly fire-resistant beforehand.

U.S. Government photos from "Shot Annie", Nevada Test Site; courtesy of Lynn Eden (Note 9)

By the late 1970s, Eden shows, the influential physicist Harold Brode of Pacific-Sierra Research focused attention on the problem of nuclear firestorms and the extent to which simultaneous ignitions were predictable effects of nuclear explosions. During the 1980s, Brode and colleagues under contract to the Defense Nuclear Agency (DNA) developed a methodology for predicting mass fire that could be incorporated into the nuclear war plans developed by the Joint Strategic Target Planning Staff (JSTPS). Their research would have made it possible to develop a fire-blast VN system that could more comprehensively predict the effects of nuclear detonations. In the early 1990s, however, JSTPS Deputy Director Vice-Admiral Michael Colley cancelled further research on nuclear fire effects. Despite Brode's innovations, Colley believed that fire was not predictable and that efforts to develop a fire-blast VN system for war planning were unnecessary: "to my mind, fire is gravy. Whatever you can get from fire just makes everything worse. And you don't need to particularly measure it. It's just there." (Note 10)

Declassified material elucidating the serious problems raised by Lynn Eden's study is scarce or otherwise highly technical. (Note 11) The following documents from the 1950s and early 1960s provide evidence of the powerful influence of a "blast damage frame" not only in the military but also among nuclear weapons experts at the State Department. Other documents show, however, that some policy experts believed that mass fires were routine effects of high-yield nuclear weapons. When the newly-created JSTPS developed the first SIOP [Single Integrated Operational Plan] in 1960, some senior officials, including White House Science Adviser George Kistiakowsky and Chief of Naval Operations Admiral Arleigh Burke argued that the JSTPS was building "overkill" into nuclear war plans by not taking fire effects into account. Despite these arguments, however, a JSTPS report shows that it believed that mass fires were unpredictable and not worth planning for. The blast damage frame held by PV and JSTPS officials remained unchallenged until the early 1980s when nuclear weapons effects experts Harold Brode convinced the Defense Nuclear Agency (DNA) to sponsor studies on nuclear fire storms. Most of the publicly available Brode material is highly technical but the document excerpted below provides evidence of his proposition that mass fires are predictable weapons effects.


Note: The following documents are in PDF format.
You will need to download and install the free Adobe Acrobat Reader to view.

Document 1: U.S. Air Force Project RAND, Implications of Large-Yield Nuclear Weapons, 10 July 1952, Top Secret, Excised Copy
Source: FOIA Release, U.S. Air Force

At the beginning of the 1950s, when thermonuclear weapons were being developed, the Air Force's civilian research arm, Project RAND (acronym for "research and development") began exploring the implications of these fearsome new weapons. The study, conducted by Bernard Brodie, Charles Hitch, and Ernst Plesset, was widely briefed in the national security establishment up to President Truman himself. Characterizing thermonuclear weapons as "killers and fantastically destructive," the RAND analysts looked at such problems as weapons effects, the impact of Soviet acquisition, the possibility of defense, and offensive uses. The discussion of weapons effects looked beyond the destruction caused by blast and straightforwardly addressed the "increased probability of fire." Looking at the impact of a five megaton bomb dropped on a city, the analysts concluded that "a firestorm must be expected as far out" as four miles from ground zero. Given such devastating effects, the authors concluded that "large-scale reciprocal use of atomic and thermonuclear weapons against cities would not fall short of national suicide for both sides." That this made nuclear war wholly inconsistent with any rational political objective would trouble Brodie and his colleagues for the rest of their careers. While the Brodie-Hitch-Plesset scenario depicted the varied effects of a thermonuclear explosion, the "blast effects frame" had strong influence at RAND and the organization's work tended to reinforce it. (Note 12)

Document 2: Memorandum for the File, "Discussion with the Secretary" by Special Assistant to the Secretary of State for Atomic Energy Gerard C. Smith, 16 March 1957, Top Secret
Source: National Archives, Record Group 59. Records of the Department of State, Records of the Deputy Assistant Secretary of State for Politico-Military Affairs, Subject Files of Special Assistant for Atomic Energy and Aerospace, 1950-1966, box 6, II.3.B Weapons Effects - 1955-57

This document on the nuclear education of Secretary of State John Foster Dulles illustrates the predominance of the "blast damage frame." After Dulles made press conference statements that underestimated the fallout hazards associated using low-yield nuclear weapons, his special assistant Gerard C. Smith took Dulles aside to give him a better grasp of the issue. In particular, Smith wanted to make sure that Dulles understood that the U.S. Air Force would never accept tactical nuclear weapons as a substitute for strategic bombardment. After Dulles took issue with the statement that megaton nuclear strikes on the Soviet Union would "cause tremendous destruction of life and property," Smith realized that his chief needed better information but he could go no further at the moment because others in the room lacked "Q" clearances for atomic energy information. When Smith found an opportunity for a briefing, Dulles learned that "if the SAC strikes were successful … Russian casualties would be in the tens of millions." The criterion that Smith used for nuclear devastation was blast destruction: if the megatonnage of nuclear weapons dropped on Soviet cities was "divided into the total area of Russia, it would "result in substantial over-pressure over the whole area." Had he included mass fire damage, casualties would have been stated as even greater.

Document 3: Letter from Captain John H. Morse, Special Assistant to the Chairman, Atomic Energy Commission, to Lewis Strauss, Chairman, Atomic Energy Commission, 14 February 1957, Secret
Source: Dwight D. Eisenhower Library, Records of Special Assistant for National Security Affairs, NSC Series, Briefing Notes Subseries, box 17, Target Systems (1957-1961)

John H. Morse, who had served as director of atomic planning in the air operations branch of the U.S. European Command, brought his experience to bear in this discussion of nuclear planning by the military. While he does not mention blast or fire effects, his description of Air Force target planning methodology is plainly relevant to the "blast damage frame," with such references as "damage to concrete structures" and the requirement for a "high probability of cratering runways." Particularly fascinating is Morse's concern over the "destructive and disruptive nature of nuclear weapons" with megaton yields: "the cumulative or ancillary effects may be as great or greater than primary damage." While Morse only mentions fallout, firestorms are consistent with his brief discussion of "cumulative" effects. Significantly, he called attention to a Pentagon study that questioned the tendency of nuclear planners to overlook "bonus" effects; the report concluded that by taking into account the total impact of a nuclear explosion, lower-yield weapons could be used to achieve the "desired destruction." Morse noted that the Pentagon "rigorously suppressed" this study and destroyed all copies. (Note 13)

Document 4: "Nuclear Weapons", Briefing for Secretary Herter, with "Weapons Effects Briefing for Undersecretary of State," 27 March 1957
Source: National Archives, Record Group 59, Records of the Department of State, Records of the Deputy Assistant Secretary of State for Politico-Military Affairs, Subject Files of Special Assistant for Atomic Energy and Aerospace, 1950-1966, box 6, II.3.B Weapons Effects - 1955-57

Not long after becoming Under Secretary of State in February 1957, Christian A. Herter received a standard briefing, apparently involving Gerard Smith and General Herbert Loper, Assistant to the Secretary of Defense for Atomic Energy, on nuclear weapons and their effects. The outline of the briefing on weapons effects provided details on the vulnerabilities of inanimate objects and humans. With respect to the vulnerabilities of buildings, etc., blast was the "primary destruction agent." The outline identified thermal radiation as a cause of fires, even noting the "ignition energies" for fuels, wood, and paper, but apparently did not see fire as important as blast in producing destruction. On the vulnerabilities of humans, the outline gave full attention to gamma radiation and fallout dangers, treating thermal radiation and blast effects as relatively less important as casualty producers. That Smith and his colleagues did not take fire into account is suggested in the section of "protective measures" where "cover" is listed as a way to evade thermal effects. As Eden's scenario suggests, however, those who might seek shelter in basements in Arlington or Capital Hill would not find protection from heat or carbon monoxide.

Document 5: Letter from Brigadier General George S. Brown, Military Assistant to Secretary of Defense, to Commander-in-Chief Strategic Air Command Thomas Powers, 23 November 1960, Top Secret
Source: FOIA Release, Department of Defense

Not everyone was persuaded by Air Force arguments favoring "overbombing", but it remained difficult to change the direction of nuclear planning. In 1960, various organizational pressures led President Eisenhower to approve the creation of a Joint Strategic Target Planning Staff (JSTPS) that through interservice coordination would produce a National Strategic Target List (NSTL) and a comprehensive nuclear war plan, the Single Integrated Operational Plan (SIOP). JSTPS's director would be the Commander-in-Chief of the Strategic Air Command (CINCSAC), thus putting SAC effectively in charge of nuclear planning. This development had been a source of heartburn for Chief of Naval Operations Arleigh Burke, who saw it as an Air Force power play to erode the JCS's authority in nuclear planning. Burke failed to convince Eisenhower that there was a serious problem and during 1960 the JSTPS developed what would become known as SIOP-62 (for fiscal year 1962). A plan for a massive strike against military, government control, and urban-industrial targets in the Soviet Union, China, and allied countries, the SIOP, as David Rosenberg has put it, produced the "institutionalization of overkill." (Note 14) The relationship of the SIOP to the "overkill" phenomenon and the "blast damage frame" is evident from General George S. Brown's letter to CINCSAC Thomas Power. What Brown wanted Power to know was that Secretary of Defense Thomas S. Gates believed that the SIOP needed more work, not least because "weapons requirements are perhaps increased unrealistically by restricting consideration of weapons effects to blast damage" while there had been no effort to "apply thermal effects and even radiation" to target planning. It is possible that Chief of Naval Operations Admiral Arleigh Burke or White House Science Adviser George Kistiakowsky (see documents 6 and 7) made Gates aware of the problem of excessive weapons requirements. However, he would have no opportunity to follow up the problem because the Eisenhower administration was then on its way out.

Document 6: Chief of Naval Operations Admiral Arleigh Burke to Flag and General Officers, "National Strategic Target List and Single Integrated Operational Plan," Special Edition Flag Officers Dope, 4 December 1960, Secret
Source: U.S. Navy Operational Archives, Arleigh Burke Papers, Memos and Ltrs (NSTL)

Arleigh Burke acquiesced in Eisenhower's decision to go forward with the JSTPS; not surprisingly he saw serious problems in SIOP-62. While characterizing it as a "good first" step in a report to the Navy's high command, he made it plain that he saw significant weaknesses. Like Secretary Gates, Burke was especially concerned about the SIOP's damage criteria which he believed were unrealistically high: "the damage on the facilities at Hiroshima which were incapable of any type of operation and never were repaired in peacetime would not have met this damage criteria." He also believed that using blast effects only to compute nuclear devastation was a drawback: "it is certain that there will be many firestorms and damage by neutrons beyond the area of blast damage" (see page 5).

Document 7: Joint Chiefs of Staff Paper 2056/208, 27 January 1961, enclosing memorandum from Secretary of Defense Gates to the Chairman, JCS, 20 January 1961, and Excerpts from Memorandum to President Eisenhower from Special Assistant to the President for Science and Technology George Kistiakowsky, 25 November 1960, Top Secret, Excised Copy
Source: National Archives, Records of the Joint Chiefs of Staff, RG 218, 3205 Target Systems (17 Aug 59) Sec. 9

Burke and Gates were not the only senior officials who expressed reservations about SIOP-62. The White House science adviser, Harvard chemistry professor George S. Kistiakowsky, traveled to SAC headquarters in November 1960 where he heard JSTPS briefings on the war plan. He shared Burke's doubts about the damage criteria because they led to "unnecessary and undesirable overkill." Kistiakowsky also recognized the limits of the blast damage approach: he wrote to Eisenhower that "The JSTPS used blast effects as the only criterion of damage and neglected thermal radiation, fires which will be caused by it and fall-out" (see p. 1914). The day he turned in his report, Kistiakowsky briefed Eisenhower on his findings; the latter confided to his naval aide Captain E.P. Aurand that the briefing "frightened the devil out of me." Eisenhower was surprised by the numbers of targets and the huge scale of the attack plan. Believing that there had to be limits, he wondered whether it might be better to let SAC "have one wack--not ten wacks" at each target. But it was too late for Eisenhower to change the direction of U.S. war planning. (Note 15)

Document 8: JSTPS Report 3-61, Staff Study, "Damage Criteria," 3 June 1961
Source: FOIA Release, U.S. Strategic Command

The huge excisions from this study show how sensitive the issue of nuclear devastation has been to the U.S. military. Certainly, the residual content that has been left shows the supremacy of the "blast damage frame" for measuring nuclear weapons destruction. Thus, Annex A begins by discussing the vulnerability number (VN) system used for predicting blast damage. That the JSTPS wholly discounted Gates', Burke's, and Kistiakowsky's thinking about the likelihood of mass fires is evident on page 10. There, the JSTPS analysts briefly discussed the problem and possibility of nuclear fire effects. After noting that the spread of fire depended on a number of considerations and reviewing the damage to Hiroshima and Nagasaki, they concluded that firestorms were unpredictable: They are "not a special characteristic of nuclear explosions. They may or may not occur."

To help make their case about the unpredictability of fire effects, the JSTPS analysts argued that the plutonium weapon that devastated Nagasaki did not produce a firestorm. Yet, strong evidence shows otherwise, despite the important geographical differences between Hiroshima (where mass fire is undisputed) and Nagasaki. Citing MIT scientist Theodore Postol, who has argued that Nagasaki was "Hiroshima in a fireplace," Eden maintains that mass fire also occurred at Nagasaki. That city is "Located in an upward sloping valley" and the valley's sides "acted like the walls of a giant fireplace and the upward slope acted like a flue." Fire spread up the flue, from low to high ground. (Note 16)

Document 9: Harold Brode and Richard D. Small, Fire Damage and Strategic Targeting, PSR [Pacific Sierra Research] Note 567, sponsored by DNA, Washington, D.C. (Los Angeles: PSR, June 1983), excerpts
Source: courtesy of Lynn Eden

Nearly twenty years after JSTPS analysts dismissed the problem of mass fire physicists Harold Brode and Richard Small were intensively researching the fire damage issue in studies for the Defense Nuclear Agency. In this document, their first major study on fire for DNA, Brode and Small tacitly disagreed with the earlier JSTPS approach by arguing that a nuclear detonation "ensures a very large number of ignitions and the rapid development of a large area fire." Therefore, in spite of such uncertainties as weather and civil defense measures, among other factors, "fire damage can be predicted with useful consistency" and can be "as reliable as corresponding blast damage predictions." Because fire was likely to cause "more complete and permanent damage," including it in damage prediction methodology would improve the accuracy of nuclear war planning. Recognizing that nuclear war is a nasty business, they wrote, "A new damage methodology including fire effects need not wait for changes in national objectives. If 'moderate damage' to 'steel frame' buildings is the appropriate guide for destroying a city of a million or more inhabitants, then fire can only complete the job more effectively." (Note 17)

Eden's book shows that Brode made significant progress in winning support for his methodology, but skepticism about the predictability of firestorms remained a powerful obstacle and the JSTPS cut off funding for additional research on the subject in the early 1990s. (Note 18) Documentation on the JSTPS's decision remains classified; so does information on the Pentagon's apparently more positive interest in the mass fire problem in the late 1990s. By shedding light on basic elements of the U.S. nuclear posture these issues warrant further investigation by journalists and scholars alike.


1. Quotation from Chief of Naval Operations Admiral Arleigh Burke, December 1960 (see document 6)

2. Eden chose a 300 kiloton weapon because "many weapons in modern arsenals have yields of 300 kilotons or more." On the U.S. side, Minuteman IIIa's and MX's have yields of 335 and 300 respectively, while Russian ICBM warheads have yields ranging from 550 to 750 kilotons. See Lynn Eden, Whole World on Fire: Organizations, Knowledge, and Nuclear Weapons Devastation (Ithaca, NY: Cornell University Press, 2004), pp. 270-271.

3. See Eden interview for "The American Experience: Race for the Superbomb," at <http://www.pbs.org/wgbh/amex/bomb/filmmore/reference/interview/eden.html>. A study done for the Defense Nuclear Agency in 1990 estimated fire damage radii as "2 to 5 times larger" than for blast effects. Eden, Whole World on Fire, p. 246.

4. In making this argument, Eden builds upon David A. Rosenberg's pioneering research on the history of U.S. nuclear strategy; see Rosenberg, "The Origins of Overkill: Nuclear Weapons and American Strategy, 1945-1960," in Steven Miller, ed., Strategy and Nuclear Deterrence (Princeton: Princeton University Press, 1984), pp. 113-182.

5. Stephen Schwartz, et al. Atomic Audit : The Costs and Consequences of U.S. Nuclear Weapons Since 1940 (Washington, D.C.: Brookings Institution, 1998).

6. This description of blast effects draws upon Federation of American Scientists, "Nuclear Weapons Blast Effects," at <http://www.fas.org/nuke/intro/nuke/blast.htm>. Quotation from Eden, Whole World on Fire, at p. 19.

7. Image from Samuel Glasstone, The Effects of Nuclear Weapons (Washington, D.C.: Government Printing Office, 1964), as reproduced, with the other stages of the air burst, on the "Trinity Atomic Web Site" <http://www.cddc.vt.edu/host/atomic/nukeffct/airburst.htm>.

8. For examples of VN ratings of targets, see Natural Resources Defense Council, The U.S. Nuclear War Plan: A Time for Change (Washington, D.C.: Natural Resources Defense Council, 2001), p. 122; see also <http://www.nrdc.org/nuclear/warplan/warplan_ch5.pdf>.

9. For a digitalized film of the same event, shot from slightly different perspective, viewable with "Quick Time," see <http://www.cddc.vt.edu/host/atomic/nukeffct/effct01a.mov> on the "Trinity Atomic Web Site."

10. Eden, Whole World on Fire, p. 275.

11. Eden worked from documents released through Freedom of Information Act requests, some of which were prepared jointly by her and the National Security Archive. She also worked from other archival documents, unclassified government reports, and interviews.

12. Fred Kaplan, The Wizards of Armageddon (Stanford, CA: Stanford University Press, 1983), pp. 81-82.

13. At first glance, it would seem that Morse was referring to the Army-Navy study "Project Budapest," that, according to David A. Rosenberg, was a "devastating critique" of Air Force nuclear planning. But "Budapest" was not presented to the Joint Chiefs until August 1957, some months later, so the study that Morse mentions must have been a forerunner. For details on "Budapest," see Rosenberg, "The Origins of Overkill," p. 161.

14. Ibid., pp. 114-115, 174. For further discussion of Burke's battle, see Kaplan, Wizards, pp. 264-266.

15. Rosenberg, "Origins of Overkill, p. 118.

16. Eden, Whole World on Fire, p. 20.

17. Ibid, pp. 233-238, for discussion of this study.

18. Ibid, chapters 9 and 10.


home | about | documents | news | publications | FOIA | research | internships | search | donate | mailing list

Contents of this website Copyright 1995-2004 National Security Archive. All rights reserved.
Terms and conditions for use of materials found on this website.