A brief and source audit for a fallout study of one small island, from the Home Office's own exercises and the British scholarship that answered them, with what the Warsaw Pact's released plans say and do not say about Britain as a target.
Thesis
Britain is the case where the fallout plume, not the fireball, is the subject, and where the state's own numbers and the scholars' numbers sit furthest apart. The Home Office ran Square Leg in September 1980 with 150 weapons and 280 megatons on the country and reported 8.5 million blast dead; Openshaw, Steadman and Greene put the same attack through a computer and got 29 million, more than half the population. Both figures are in the record. So is the 1955 Strath report, which found twenty-five years earlier that ten hydrogen bombs would kill twelve million and that fallout "presents problems of a revolutionary character," and was classified until 2002. The study is that disagreement on the map: the plumes drawn over real counties under the winds the Home Office assumed, the regional bunkers marked where they were, and the readout carrying the government's estimate and the scholars' side by side.
The Soviet side is the withheld tier made literal. The Warsaw Pact's 1979 Seven Days to the River Rhine, released by Poland in 2005, draws nuclear strikes across West Germany, the Low Countries, Denmark and Italy and leaves Britain and France unstruck, apparently because both could answer in kind. Whether Britain would have been spared in the general war that Square Leg assumed, or struck as Square Leg's planners assumed, is exactly the question the two sides' documents disagree on.
What the record holds
The British studies
| Element | Figure | Source and tier |
|---|---|---|
| Strath report, 1955 | Ten hydrogen bombs of 10 Mt at night without preparation: 16 million casualties, up to 12 million dead (nine million blast and heat, under three million radiation), four million seriously injured, 13 million needing shelter for a week; London alone four million casualties; classified until April 2002 | Hennessy, The Secret State (2010), pp. 167–172; the report at The National Archives. Documented |
| Square Leg, 11 to 25 September 1980, strike on the 19th | 150 weapons of 500 kt to 3 Mt, 280.5 Mt in all, ground, air and underwater bursts; partial bomb plots released for England, Wales and Scotland; inner London not struck; the Home Office's own note that a real attack would exceed 1,000 Mt | Campbell, War Plan UK (1982); the regional war diaries at The National Archives. Documented, with the bomb plot partly withheld |
| Square Leg casualties, official 1982 estimate | 8.5 million blast dead, 2.5 million radiation dead, 2 million severely injured, 41 million uninjured | Home Office, via Campbell. Documented |
| Square Leg casualties, independent estimate | 29 million dead (53 percent), 7 million seriously injured (12 percent), 19 million uninjured (35 percent) | Openshaw, Steadman and Greene, Doomsday: Britain after Nuclear Attack (1983). Documented |
| Hard Rock, 1982 | The successor exercise, cancelled when councils refused to take part; its scenario called the campaign "Public, Do-It-Yourself Civil Defence" | Campbell; Wikipedia, Protect and Survive. Documented as a cancellation |
| Protect and Survive | Produced 1974 to 1976, 2,000 copies issued secretly in 1976, published in May 1980 at 50 pence; Butler's 1981 paper found strict adherence to the stay-at-home advice would add "in the region of 12 million" deaths in a countervalue attack | The pamphlet and films; Wikipedia, Protect and Survive, citing the Home Office F6 papers. Documented |
| The JIC's target assessment | The Joint Intelligence Committee's 1960s papers on the Soviet threat to the United Kingdom, with the assumed targets by category, released at The National Archives | Hennessy, The Secret State, chapter on the "probable nuclear targets"; to be read before any of it becomes data. Documented, figures to be checked |
| Regional government | The eleven regions and their headquarters in the final configuration: Cultybraggan, Hexham, Skendleby and Loughborough, Bawburgh and Hertford, Kelvedon Hatch, Crowborough, Bolt Head and Chilmark, Brackla and Wrexham, Drakelow and Swynnerton, Hack Green and Goosnargh, Ballymena; the network exposed in 1963 | Wikipedia, Regional Seats of Government; Campbell. Documented |
| Fallout monitoring | The UKWMO and the Royal Observer Corps posts, about 870 of them, the only instrument the state had for reading the plumes it planned for | Wood, Attack Warning Red (1976); the ROC record. Documented |
The Soviet side
| Element | Figure | Source and tier |
|---|---|---|
| Seven Days to the River Rhine, 1979 | A NATO first strike on the Vistula answered by Soviet strikes on West Germany, Belgium, the Netherlands, Denmark and north-east Italy, about 7.5 Mt in all, Vienna with two 500 kt weapons, Vicenza, Verona and Padua one each; Britain and France excluded | Released by Sikorski in 2005; Wikipedia, Seven Days to the River Rhine. Documented |
| SS-20 | 435 missiles at 48 sites by 1986, three 150 kt warheads each, 5,000 km, the weapon Britain expected; withdrawn under the INF treaty | Wikipedia, RSD-10 Pioneer. Documented in total; the share aimed at Britain is not in the record |
| What Britain assumed | Square Leg's planners struck Britain with 150 weapons in a general war; the released Pact plan struck the continent and not Britain in a theatre war. The two are not the same war, and the study must say which it is enacting | The withheld tier, drawn as such |
British forces
| Element | Figure | Source and tier |
|---|---|---|
| Polaris | Four Resolution-class boats at Faslane, 16 missiles each, A3T with three 200 kt warheads, Chevaline from mid-1982 on Renown and on the last boat by 1987; one boat always in refit and one always at sea | Wikipedia, UGM-27 Polaris. Documented |
| American forces in Britain | The F-111s at Lakenheath and Upper Heyford, the SSBN tender at Holy Loch, the cruise missiles at Greenham Common from November 1983 and Molesworth, the RAF's own WE.177 force | Databook vol. 1; the RAF record. Documented; the target value of each is the JIC's, not the study's |
What the study would show
- Square Leg's bomb plot as far as it was released, drawn as documented; the withheld inner-London part as a black bar over the capital, the study's first use of the withheld tier for a whole city.
- The plumes under the exercise's own wind assumptions, from the fallout lab's Glasstone contours, over the HYDE 1983 grid that is already prepared, with the ROC posts as the places the state would have read them from. The readout gives three numbers where the others give two: the Home Office estimate, Openshaw's, and the engine's own sum, with the difference explained by the shelter and evacuation assumptions each made.
- The regional headquarters as sites, with the bunkers that were themselves struck in the exercise marked so.
- Strath as a second act from 1955: ten 10 Mt weapons on the 1950 grid, the state's first look at the hydrogen bomb, with its twelve million against the engine's sum.
- The Soviet plan as a question: the Seven Days map with Britain blank, next to Square Leg's Britain, and the reasons each side had.
Built, 9 September 2026
#/study/britain-80 plays Square Leg's attack and #/study/britain-80/strath the ten bombs of 1955. The sites are built by scripts/build-britain-1980.py: 57 sites in the categories the record names, the airfields, naval bases, ports and command sites of 1980, all drawn withheld because the plot that may or may not have included them is; the seventeen regional government headquarters of the final configuration; and the 29 ROC group controls where the fallout would have been read. The bomb plot itself could not be had: the Ministry refused the requests of 2009 and 2011 and the written answer of December 1980 declined to name targets, so the weapons fall by a stated rule from the documented totals, Campbell's 69 ground and 62 air bursts, 1 Mt on the ground and 2 Mt in the air inside the documented range, the cities taken as the most populous cells of the 1983 grid with inner London left out as the exercise left it. The timing is the exercise's own, the first strike from 12:01 to 12:10 and the second drifting in from 13:00 to 15:00, and so is the southerly wind.
| Square Leg on the engine, 48 hours in | Engine, GHSL 1985, union | Engine, GHSL 1985, summed per target | Engine, HYDE 1983, summed | Home Office, 1982 | Openshaw, Steadman and Greene, 1983 |
|---|---|---|---|---|---|
| Blast dead | 15 million by the 1961 method | 16 million | 14 million | 8.5 million | 29 million dead in all |
| With mass fire | 30 million, the Postol bound | 36 million | 34 million | ||
| Fallout | 2.6 million acute deaths among the blast and fire survivors, 46 million under the plumes | 1.4 million, 57 plumes summed | 8 million | 2.5 million radiation dead | |
| Blast, fire and fallout in sequence | 33 million | ||||
| Injured by blast | 14 million | 23 million | 23 million | 2 million | 7 million |
The union column is the one to read: every grid sample counted once, in the most severe band any detonation reaches it with, then fire as Postol's bound, then fallout among those left alive, and the last row all three in sequence. The summed columns show what overlap does: six million of the fire figure were people under two or more fire zones. The fallout figure rises from the per-plume sum because the union takes the highest dose any plume gives a sample where the sum took each plume's contour at its own dose, and it now counts only the living.
The two grids agree on the blast and disagree on the fallout by a factor of six, and the reason is the shape of a plume: tens of kilometres wide and hundreds long, it falls between HYDE's nine-kilometre cells and takes whole cells' people with it, while the 30-arc-second grid puts the people where they were. The finer grid is the one to read; the coarser column stays to show what resolution does to a fallout sum.
The engine brackets both estimates rather than choosing between them, and the readout says why: the Home Office assumed its shelters, Openshaw modelled the country without them, and the engine assumes none and sums every plume separately, so the people under more than one plume are counted more than once. The plumes themselves are the picture the Home Office did not publish: 57 ground bursts on a southerly wind put the whole island north of the Thames under fallout by the second day, with the regional headquarters at Hexham, Cultybraggan and Goosnargh sitting inside it.
Strath drops ten 10 Mt ground bursts at night on the ten most populous cells of 1950 from Long Range Aviation's fields. On the 1950 grid the engine reads 5.9 million dead by the blast-only method and 19 million with mass fire against the report's twelve million, which was blast, heat and radiation together; the report's own targets are not in the public summary.
What it must not claim
- Square Leg's full bomb plot is not released. The study draws what was, marks what was not, and does not invent the rest.
- No Soviet target list for Britain is in the record. The SS-20 total is documented; its allocation is not, and the Pact's released plan is for a different war.
- The casualty disagreement is the subject, not a thing to resolve: the Home Office, Openshaw and the engine each made different assumptions about shelter, evacuation and fire, and the readout must say what each assumed.
- The JIC target papers must be read at the Archives or in Hennessy before any target is drawn from them.
Data needed
- Square Leg's released bomb plot: target, yield, burst height, from Campbell's tables and the National Archives files.
- The exercise's wind assumptions for the plumes, from the same files.
- The regional headquarters and ROC posts as positions.
- Strath's ten targets, if the report names them; otherwise the ten largest 1950 cities as a stated rule.
- HYDE 1983 (prepared, committed) and 1950 (prepared).