Charles F. Carter, Jr., Colonel, U.S. Army, Chief Research and Analysis Division DASA/DODDAC [Department of Defense Damage Assessment Center] to Captain F. V. BENNETT, USN, Military Assistant to ATSD [Assistant to the Secretary of Defense] (AE), "AEC/DOD Study," 28 August 1962, Secret, excised copy
National Security Archive
A 1962 secret memo reveals how the U.S. meticulously modeled civilian casualties for a hypothetical Soviet nuclear war, exposing the grim calculus behind the era’s ‘clean‑nuke’ rhetoric.
Source: Charles F. Carter, Jr., Colonel, U.S. Army, Chief Research and Analysis Division DASA/DODDAC [Department of Defense Damage Assessment Center] to Captain F. V. BENNETT, USN, Military Assistant to ATSD [Assistant to the Secretary of Defense] (AE), "AEC/DOD Study," 28 August 1962, Secret, excised copy Date: Aug 28, 1962 Archive: Record Group 330, Records of the Department of Defense, Office of the Assistant to the Secretary of Defense (Atomic Energy), Accession 69-A-2243, "AW- Ecological Study, Volumes I and II" Collection: "Clean" Nukes and the Ecology of Nuclear War Aug 30, 2017
Editorial Analysis
Original analysis by the DriftSeas editorial desk. The complete primary-source document, transcribed from the National Security Archive scan, appears in full below.
A Secret Blueprint for the Aftermath of a ‘Clean’ Nuclear War
On 28 August 1962 Colonel Charles F. Carter, Jr., chief of the Defense Atomic Support Agency’s Research and Analysis Division, sent a status memorandum to Captain F. V. Bennett, the military assistant to the Assistant Secretary of Defense for Atomic Energy. The memo was a routine‑sounding update on a joint Atomic Energy Commission–Department of Defense (AEC/DOD) study of the ecological and biological consequences of a large‑scale nuclear exchange with the Soviet Union. What the document reveals, however, is far more than bureaucratic delay: it is a window onto the Cold War’s most chilling calculus – the systematic modeling of civilian casualties under the assumption that future weapons could be “cleaner,” i.e., produce less fallout.
The memo was produced at a moment when the United States was finalising its “New Look” nuclear strategy, which emphasized massive retaliation and the development of high‑yield, high‑precision weapons. The AEC/DOD study, initiated in early 1962, was tasked with quantifying how many Soviet civilians would survive a series of hypothetical attacks ranging from 1,000 to 10,000 megatons. The study’s parameters—specific yields (1 MT, 5 MT, 20 MT), target sets (military installations, missile factories, major industrial centres), and even wind‑speed assumptions—were deliberately chosen to generate worst‑case casualty estimates while also allowing the analysts to compare “clean” versus “standard” weapons.
The Numbers Behind the Horror
Carter’s memorandum lists three attack scenarios. The smallest, a 1,000‑megaton strike, would employ mostly 1‑megaton weapons and aim for a 50 % probability of moderate damage to key Soviet sites. The middle tier, 3,000 megaton, mixes in more 5‑ and 20‑megaton warheads to raise the probability of severe damage, while the largest, 10,000 megaton, adds a substantial number of 20‑megaton weapons and expands the target list by 180 sites to push the probability of catastrophic damage above 90 %.
The study’s methodology is laid out in painstaking detail: the USSR is divided into roughly 7,500 grid cells; a stabilized radioactive cloud is modeled; fallout deposition is calculated using particle fall velocity, wind speed, and shear; and a “residual number” (0.005–0.70) gauges shielding effectiveness. The analysts even fixed the circular error probable (CEP) of all delivery systems at 3,000 ft, sidestepping the messy reality of varying accuracy. By choosing a mean spring wind that blows west‑to‑east, the model artificially simplifies atmospheric transport, yet the resulting casualty tables would have fed directly into strategic planning documents.
What the Delay Tells Us
Carter’s memo is unusually candid about the project’s slippage. He cites “complexity of the study,” “changes in parametric inputs,” and “computer programming and system design complexities” as reasons for missing earlier deadlines. In 1962, the DoD was still wrestling with the first generation of large‑scale digital computers; the sheer volume of data—population grids, soil composition, weapon yields—required custom code that was, by today’s standards, primitive. The fact that the Department of Defense Damage Assessment Center (DODDAC) needed an extra week to deliver its findings underscores how novel and technically demanding it was to model a nuclear war’s ecological fallout.
The memo also notes that Hal Hollister of the AEC was formally informed of the delay on 27 August, a day before the memorandum was sent. This procedural detail hints at the inter‑agency tension that often accompanied joint studies: the AEC, responsible for weapon design, and the DoD, responsible for target selection and strategic doctrine, were each trying to protect their own institutional interests while feeding a shared narrative of deterrence.
Legacy of the ‘Clean‑Nuke’ Calculus
Although the document was classified for decades, its declassification in 2017 shows that the United States spent the early 1960s not only building bigger bombs but also rigorously quantifying their humanitarian impact. The study’s focus on “clean” weapons—those that would produce less radioactive fallout—reveals a paradox: even as policymakers touted the moral advantage of reduced fallout, the underlying models assumed a total annihilation of millions of civilians.
The significance of Carter’s memorandum lies in its illustration of how scientific modeling was weaponized during the Cold War. The same grids and residual numbers that fed into this 1962 study later informed civil‑defense pamphlets, public‑information campaigns, and the very rhetoric that justified the doctrine of massive retaliation. Today, as renewed great‑power competition raises the specter of high‑yield weapons once more, the memo serves as a reminder that the technical details of fallout modeling are inseparable from the moral calculus of nuclear policy.
Why It Still Matters
For historians, the memo fills a gap between the abstract strategic doctrines of the Kennedy administration and the concrete, data‑driven analyses that underpinned them. It shows that the United States was already attempting to predict, in granular detail, the civilian toll of a nuclear exchange—information that would later be used both to deter adversaries and to reassure the American public that “clean” weapons were a humane alternative. In an era where nuclear arms control debates increasingly focus on low‑yield, “tactical” weapons, revisiting the 1962 AEC/DOD study reminds us that any claim of a “clean” nuclear war is built on a foundation of grim statistical modeling, not on any genuine humanitarian benefit.
[DECLASSIFIED Authority 44472]
DASARZ 930 28 August 1962
MEMORANDUM FOR: CAPTAIN W. F. V. BENNETT, USN MILITARY ASSISTANT TO ATSD (AE)
SUBJECT: AEC/DOD Study
Reference is made to ATSD (AE) Memorandum to Chief, Defense Atomic Support Agency (DASA), subject, "AEC/DOD Study on Short/Long Term Effects of Nuclear War," dated 14 August 1962.
The purpose of this memorandum is to provide a status report of the progress of the AEC/DOD Study contribution by DASA to the fall-out and ecological study now under preparation by the AEC. The completion date of the DASA analytical effort and the compilation of data into a final report to the AEC is now estimated to be 4 September 1962. This is also the delivery date for the presentation of material requested by ATSD (AE) (reference 1).
It is recognized that there have been several inaccurately forecasted completion dates for the DASA effort. The delays in presentation of the final material are attributed to:
a. The complexity of the study project.
b. Changes in parametric inputs during the conduct of the study project.
c. The necessity for the collection and application of detailed, current and accurate data.
d. The computer programming and system design complexities and efforts commensurate with the level of detailed analysis required.
The complexity of the DASA contribution to the study has exceeded all past experience in this area. Although it is disappointing for all concerned that the estimated completion dates have not been met, it is sincerely felt that the nature and importance of the project warrant the time and effort required to accomplish the project in a thorough, competent and detailed manner.
[Declassified Authority: 44472 By: Monica Oyola-Coeur Date: 03-09-2017]
NW#:44472 DocId:32586109
DECLASSIFIED Authority 44472
DASARZ 930 SUBJECT: AEC/DOD Study
- Mr. Hal Hollister, AEC, was informed on 27 August 1962 of the delay in submission of the DASA material.
[Charles Carter Jr.] CHARLES F. CARTER, JR., LTCOL, USA Chief, Research and Analysis Division DASA/DODDAC
Copy furnished: Mr. Hal Hollister, AEC
NW#: 44472 DocId: 32586109
DECLASSIFIED Authority 44472 SECRET PARAMETERS, ASSUMPTIONS, AND CONCLUSIONS AEC/DOD STUDY
GENERAL
A study has been prepared by the Department of Defense Damage Assessment Center (DODDAC), an activity of the Defense Atomic Support Agency (DASA), as a contribution to a long term biological/ecological study under preparation jointly with the Atomic Energy Commission.
A portion of the DASA/DODDAC study addresses the relative effects on population resulting from the employment of clean nuclear weapons as compared with normal, or standard, weapons. The parameters, assumptions, discussion, and conclusions from this portion of the study are as follows:
PARAMETERS
A range of nuclear weapon attacks totaling approximately one thousand, three thousand, and ten thousand megatons was postulated against combined military and industrial targets in the USSR. A subset attack against the military targets only was also postulated but with correspondingly less yield. Discrete single weapon yields of 1MT, 5MT, and 20MT were utilized in the attacks.
Two real world target systems, based on current intelligence, were attacked hypothetically as follows:
a. Combined attack against
(1) Important military targets such as long and medium range airfields, missile facilities, nuclear weapon storage sites, submarine bases, command centers, supply depots, and others.
SECRET NW#: 44472 DocId: 32586109
DECLASSIFIED
Authority 44472
SECRET
(2) War support industries such as missile factories, nuclear
weapon production plants, atomic energy installations, aircraft plants,
large power plants, and others.
(3) Major industrial centers.
b. Military targets only as a subset of the combined attack
described above.
FOIA (b) (3) - 42 USC 2162(a) - RD DOE EO13526 6.2(a)
- In order to describe shielding of population, residual numbers, a measure of attenuation of radioactivity, were used. The residual numbers ranged from .005 to .70; the lower the residual number, the more efficient the shielding.
- Specific weapon delivery systems were not selected; accordingly, a 3000 ft. CEP was used for all weapons delivered. No other operational factors were considered.
2
SECRET
NW#:44472 DocId:32586109
DECLASSIFIED Authority 44472 SECRET ASSUMPTIONS 9. In order to avoid the extreme and erratic winter and summer winds in the USSR, a mean spring seasonal wind was chosen. This resulted in a generally west-to-east wind. 1C. Soil composition was assumed to be a meld of the two most dominant types in the USSR; however, recognition must be given to the fact that soil composition varies materially throughout the USSR. FOIA(b)(3) - 42 USC 2162(a) - RD DOE EO13526 6.2(a) DISCUSSION 12. The basic model for the calculation of casualties is a mathematical representation of the spatial distribution of the stabilized cloud's radio-activity which is then deposited downwind as a function of particle fall velocity, wind speed, and wind shear. The USSR is divided into some 7500 cells (1/2°longitude, 1/3°latitude), and assessments are made against the urban and rural population therein based upon the radioactive particles deposited. 13. The general attack philosophies are as follows: a. In the 1000 MT attack, mostly 1MT weapons were used (519 @ 1MT, 91 @ 5MT, and no 20MT) to achieve on the order of 50% probability of moderate damage to the targets selected. b. In the 3000 MT attack, the numbers of weapons by type were changed (369 @ 1MT, 141 @ 5MT, and 97 @ 20MT) to achieve a high probability 3 SECRET NW#:44472 DocId:32586109
DECLASSIFIED Authority 44472
SECRET
of severe damage. No increase in number of targets attacked was made over the number in the 1000 MT attack.
c. In the 10000 MT attack, the numbers of weapons by type were again changed (220 @ 1MT, 104 @ 5MT, and 463 @ 20MT) in order to achieve a very high probability (90% or greater) of severe structural damage where required. One hundred and eighty new targets were added in a manner to main- tain the military and non-military target symmetry.
FOIA(b)(3) - 42 USC 2162(a) - RD DOE E013526 6.2(a)
The tabular presentation of fatal casualties is shown at ____________________, and the graphical presentations are shown at ____________________ (combined attack) and at ____________________ (military target attack only).
CONCLUSIONS
Attacks employing all surface-burst normal weapons against a broad target array cause 50% (10000 MT attack) to 100% (971 MT attack) more fatalities than similar attacks employing surface-burst clean weapons.
Air-bursting of weapons is the most sparing of the population, although at about 1000 MT weight of attack, fatalities are about the same from (a) an air burst-combined targets array, (b) a clean weapon-surface burst-combined targets array, or (c) a normal weapon-surface burst attack against military targets only. For lower weights of attack, clean surface bursts will cause somewhat fewer casualties than the same air burst technique.
4
SECRET
NW#:44472 DocId:32586109
DECLASSIFIED Authority 44472 SECRET FOIA(b)(3) - 42 USC 2162(a) - RD DOE EO13526 6.2(a) 13. The targeting philosophy is as important as weapon type in affect- ing population. Attacks employing all clean or all normal weapons, as well as attacks employing all air bursts or all surface bursts, are not the most efficient and are thus not realistic. Optimum targeting involves a mixture of air and surface bursts; accordingly, fatal casualties for an actual attack should be an interpolation between the air burst and surface burst curves according to the optimum air/surface burst weapon mix. 5 SECRET NW#: 44472 DocId: 32586109
NATIONAL SECURITY ARCHIVE
National Security Archive, Suite 701, Gelman Library, The George Washington University, 2130 H Street, NW, Washington, D.C., 20037, Phone: 202/994-7000, Fax: 202/994-7005, nsarchiv@gwu.edu