Credible nuclear deterrence effects, debunking dogmatic "disarm or be annihilated" enemy propaganda. Realistic effects and credible nuclear weapon capabilities for deterring or stopping aggressive invasions and attacks which could escalate into major conventional or nuclear wars.

Monday, January 22, 2018

The January 1955 secret Fallout symposium of the AFSWP, the first analysis of the detailed data on fallout from Operation Castle, is now available on the internet archive, thanks to the Nuclear Testing Archive. UPDATED 30 Jan 2018 with declassified neutron bomb report LA-9004 on Opennet and 1957 RAND Fallout Symposium

Thank you to Martha DeMarre of the Nuclear Testing Archive, Mission Support and Test Services (MSTS), Contractor for the Nevada National Security Site of Uncle Sam, for today emailing me a scan in two parts of the terrific (formerly) secret January 1955 AFSWP Fall-Out Symposium, U.S. Armed Forces Special Weapons project report AFSWP-895, which I've put on internet archive (link is here).  This is the first major analysis (566 pages in its declassified form) of data from Operation Castle on fallout, the first major fallout hazard experience to be documented in great detail in 1954! The report was listed but a PDF was not previously available on the U.S. Department of Energy Opennet site (which currently highlights Edward Teller's dismissal of secrecy delusions in the PDF linked here).  Secrecy is damaging, as mentioned in the previous post on this blog, because it keeps the public uninformed of the key technical arguments that underpin scientific controversies, allowing abusive propaganda from bigoted, ranting communist lunatics to become "mainstream dogma", accepted by deluded, elitist pseudo-socialists as occurred after Darwin's cousin Sir Francis Galton used "authority" to push eugenics pseudo-science with a pipe dream camouflage of utopia.

AFSWP895: January 1955 Fallout Symposium, secret, front cover.

AFSWP 895: fallout dose rates at 1 hour after the 14.8 megatons surface burst Castle Bravo across Bikini Atoll, 1 March 1954.  BEWARE OF THE MAP SCALE LABELLED "NAUTICAL MILES": this map and others in the series, reproduced in the 1956 weapon test report WT-915 and then in the fallout patterns compendium DASA-1251, exaggerates the size of Bikini Atoll considerably - by a factor of about 1.5 (the East-West length of Bikini Atoll is about 33 nautical miles on the graph above, as contrasted to a reality of just 22 nautical miles (see the accurate Holves and Narver below) - and needs correction (as we have pointed out in previous posts concerning DASA-1251 and WT-915).  A good scan of an original printing of WT-915 is located here.
Bikini Atoll map with accurate scale in Nautical Miles: the East-West length is about 22 nautical miles, as contrasted to the inaccurate USNRDL map scales which give a width of about 33 nautical miles, 50% too much!  (Thus, fallout areas in Bikini Atoll are exaggerated by 1.5 squared, a factor of 2.25.)  The inaccurate maps were proliferated in other fallout reports that "compiled" inaccurate data together without checking the scales (e.g., DASA-1251, the fallout patterns compendium).  As we reported in an earlier post, this had disastrous consequences for one computer prediction method, which was sold on the basis that it reliably reproduced the false Castle-3 shot pattern (the version with the inaccurate distance scale, leading to more than a doubling of areas). 

AFSWP 895: fallout outdoor unshielded dose rates and doses after 14.8 megaton Castle Bravo across Bikini Atoll (ground zero is the reef to the immediate West of "Charlie" Island).  Upper number is dose rate, lower is accumulated dose from fallout arrival time to infinity, outdoors and without any shielding such as buildings or other shelter.

AFSWP 895 fractionation of Sr89 and Ce144 as function of fallout particle diameter in Operation Castle shot Bravo.  Compare to the fractionation data from the 1956 Redwing tests, compiled by Dr Carl F. Miller in USNRDL466.

The report also contains new photos of the fireball and cloud from the 13.5 megaton Yankee shot of Operation Castle, taken from an RB-36, including the times of each photo (which is very useful, because it shows you the evolution of the fireball into the mushroom), at pages 91-110.  On pages 110-121 there is an excellent summary of the fallout study results of the Nevada 1.2 kiloton surface burst and shallow subsurface (earth penetrator warhead simulation) bursts Sugar and Uncle, respectively, from 1951, including photos of the differences in the nature of the fallout, comparing this data to photos of fallout from the 1952 Ivy-Mike surface burst of 10.4 megatons at Eniwetok Atoll.

On pages 123-138 there is a nice paper by Dr Carl F. Miller, called "Physical and Chemical Nature of the Contaminant: Interpretation of Castle Observations", giving the fallout deposited mass per unit area for specific unit-time radiation dose rates, the averaged gamma ray energy, graphs of decay rates, and a detailed table of fallout solubility (ionic fraction of radioactivity when the fallout is mixed with water), comparing land and surface tests of Operation Castle.

(Compare this fallout solubility data to the later USNRDL reports WT-917 and WT-918.  Note that 1958 Hardtack tests report WT-1625 on page 13 briefly interprets and summarises the solubility data from Castle in WT-917 and from Redwing in WT-1317 (the WT-1317 pdf file held on the Opennet database is corrupted and will not open, but we uploaded the full WT-1317 report to internet archive, linked here, before this occurred): the land surface bursts of Castle gave 5% fallout solubility, compared to 58-73% solubility for the water surface barge bursts, whereas the Redwing effective land bursts Zuni and Tewa gave 5-25% and 8-18% solubility, respectively, using rainwater and sea water. (These percentages don't apply to individual nuclides, since the soluble fraction mainly consists volatile decay chain nuclides like I, Sr, Cs, etc., which coat the outer surfaces of fallout particles; whereas the insoluble activity is mainly refractory nuclides that condense in the inside of molten particles, like Zr, Mo, U, Pu, etc. The overall percentage of solubility is therefore the average solubility of gamma emitters, which varies with time as the fallout mixture decays, and the relative percentage of activity coming from soluble nuclides rather than insoluble nuclides, evolves.)

On pages 139-153 there is an interesting paper by Dr Chris S. Cook, called "Radiological Nature of the Contaminant: Source Gamma Energy Spectra", giving data on the fallout gamma ray spectra determined using a sodium iodide scintillation crystal and a photomultiplier tube (the scintillation or flash brightness is proportional to the energy of the gamma ray, so with a pulse height discriminator circuit you can determine the spectrum).  This is vital because the penetrating power of the gamma rays from fallout determines the protective factor of a fallout shelter, and the production of low energy gamma emitters in fallout, particularly neptunium-239 and uranium-237 (produced by the capture of a high energy neutron, above about 1 MeV, by U-238, followed by the ejection of two neutrons, i.e. a so-called n,2n reaction) reduces the danger in the fallout sheltering period of 1-14 days after a dirty bomb (with a uranium jacket on the fusion stage).  Cook reports on page 139:

"Prior to 10 days following the detonation, a large fraction of the radiations are concentrated in the vicinity of 100 kev [0.1 Mev]".

This approximately 0.1 Mev radiation is the neutron activated U-237 and Np-239 (the time of peak percentage contribution of a nuclide to T^{1.2} fallout decay is equal to the half life multiplied by 1.2/ln 2 which is a multiplication factor of 1.44).  The best data available from Castle on this was from Union, shot 4, a water surface burst.  However, excellent gamma spectrum data was obtained from land surface burst Zuni in 1956, reported in WT-1317 and related papers like USNRDL-TR-146, Spectrometric Analysis of Gamma Radiation from Fallout from Operation Redwing, which was discussed on page 19 of our Nuclear Weapons - Collateral Damage Exaggerations report.  Miller gives an excellent compilation of neutron capture to fission ratios for nuclear tests up to 1960 in tables 4 and 6 of USNRDL466, although the numbers are deleted from that table in the declassified document, so you have to instead fill in the table spaces by calculating the capture atom/fission ratios using the ratios of the dose rates in gives in table 11; for example Jangle S gave 0.106/0.1799 = 0.59 atom of U-239 per fission.  Although Navajo and Flathead are deleted from that table, the capture atoms to fission ratios are reported for those shots in other reports, when you look carefully. One piece of data is given by the declassified WT-1317 e.g. the data in Table 3.14 on page 65 states that Flathead produced 0.41 atoms of Np-239 per fission, and more data is in the declassified NV0110837.  The U239 and Np239 capture-to-fission ratios of Redwing thermonuclear weapons 3.8Mt 50% fission Cherokee, 3.53 Mt 15% fission Zuni, and 4.5 Mt 5% Navajo are reported respectively to to be 0.500, 0.427 and 0.125 respectively, on page 12 of WT-1315, shown below:

There is earlier Upshot-Knothole nuclear test fallout data on average gamma ray energy in WT-814, based on the measurement of the attenuation of gamma rays by shields of varying thickness, rather than by gamma spectrometry (the electronics needed to discriminate energy intervals from sodium iodine crystal scintillation photomultiplier pulse heights were being developed in the early 1950s).

AFSWP 895: fractionation of Sr89 Ba140 and Mo99 as function of fallout particle diameter in Operation Castle.  Note that Mo-99 is normally unfractionated since it is refractory (has a high melting point), whereas the gaseous precursors in the decay chains of strontium and barium make them effectively volatile, so they don't condense very effectively on fast-falling particles of early fallout.  This graph gives data from samples collected at 18.5 statute miles from ground zero (97,730 feet).  (There is a history of fractionation data collection at nuclear tests on pages 17-19 of Hardtack report WT-1625, other versions of which - with slightly different data deleted in delassification - are located here and here.)

AFSWP 895: measured percentage of fallout radioactivity deposited within 24 hours as a function of scaled nuclear burst altitude. The scaling procedure is to divide the actual height of burst into the cube-root of the weapon yield (i.e. 10 for 1000 kilotons).

AFSWP 895: speed of rotation of radioactive torus or toroidal circulation inside rising fireball from a 30 kiloton nuclear weapon at 1 minute, taken from Dr Kellogg's presentation (he gave an unclassified version, omitting this data on the measured speeds in the vortex, to the unclassified May 1957 congressional hearings on The Nature of Radioactive Fallout and Its Effects on Man).
AFSWP 895: Dr Kellogg's illustration showing why the cloud top heights were inaccurately measured and reported in early H-bomb tests like Mike, whose height was originally wrongly reported as 25 miles not 20 miles, due to horizontal projections from the edge being confused for the top of the cloud.

AFSWP 895: fallout from 1953 Nevada nuclear test Badger of Operation Upshot Knothole showing paths of fallout at different altitudes in the mushroom cloud: because the winds have different speeds and directions at different altitudes, there the cloud separates accordingly and fallout is distributed over a larger area than would be the case without this wind shear.  This diffusion of fallout spreads the same total amount of radioactivity over a greater area, reducing doses and dose rates to lower levels than simplistic predictions (the classic cigar shaped fallout pattern) indicate.
AFSWP 895: fallout distribution in the mushroom head and in the stem of the cloud as used in the US Army Signal Corps fallout prediction method.  Note that 90% is assumed to be in the mushroom head, and that 10% is in the stem (at lower altitudes), but the average size of the particles in the stem are larger than those in the mushroom head.  This type of analysis, based on trying to reconcile theory with observed fallout data, is the source of the statement about the assumed distribution in Glasstone's Effects of Nuclear Weapons.
AFSWP 895: one effort (by Lt Col Lulejian) to reconstruct the fallout distribution across Rongelap Atoll in the 15 megatons Castle Bravo test on 1 March 1954 based on wind data analysis.  This is controversial and possibly very unreliable due to the inclusion of Eniwetok Atoll data (200 miles to the West of Bikini, i.e. 200 miles upwind!).  However, it shows that efforts were being made to try to determine the whole fallout pattern for the January 1955 Fall-Out Symposium.
AFSWP 895: Lulejian's effort to model fallout distribution doses to 48 hours across Rongelap Atoll in the 15 megatons Castle Bravo test on 1 March 1954 based on wind data analysis, combined with radiation measurements made on atoll islands.   The only Castle test where the entire fallout pattern was measured was 13.5 megaton Yankee, using ships and aircraft to survey the ocean and then to correct the measurements for the large protective factor of the water (when the fallout hits the water, most of the activity, whether soluble or in micron sized insoluble metallic particles within the relatively large calcium hydroxide flakes, ends up dispersed within the 100 metre thick surface water above the thermocline, attenuating the surface dose rate to something on the order of 500-1000 times less than the dose rate you get when the same amount of fallout is deposited on a land surface).  1956 Redwing nuclear tests showed that water surface bursts like Yankee in the 80% humidity air of Bikini atoll produce similar local fallout distributions to land surface bursts, and Yankee probably gave a very similar fallout distribution to Bravo's shot time wind fallout pattern.  This is similar to RAND Corporation's analysis of the Bravo fallout.  AFSWP 895 also gives Schuert's elaborate and misleading Bravo fallout reconstruction (later reprinted in USNRDL report WT-915), which puts too much activity in the highest dose rate contours, violating the area versus dose rate plots given by four land and water shots in Redwing, when scaled to 1 fission megaton (see WT-1316, Figure 2.45).  (Also, see Kelloggs testimony on page 105 of the 1957 congressional hearing Nature of Radioactive Fallout and Its Effects on Man, where Kellogg notes that the percentages of local fallout for land and water surface bursts in Redwing were actually very similar, and an earlier analysis to the contrary ignored Na-24 and use the wrong conversion factor between dose rate and activity; unfortunately the corrected data was ignored and the earlier mistaken analysis is quoted by Glasstone and Dolan 1977, and is also quoted by Chuck Hansen in his 1988 book US Nuclear Weapons.  To summarise, the initial analysis of the Flathead and Navajo water surface burst tests of Redwing indicated only about 30% of activity down in 24 hours, but the reanalysis by B. L. Tucker of RAND Corp, allowing for Na-24 and the correct dose rate to fissions conversion factor, gave 65-70%, which is within the error limits on land burst data.  The actual percentage refers to effective gamma dose rates not specific nuclides; refractory nuclides are concentrated on large particles which arrive in local fallout, while volatile nuclides that condense at late times on the remaining very small particles in the cloud, mostly come down later on more distant fallout.)  For Yankee's dose rate versus area data, see table here. (This was discussed in previous posts on this blog.)  There is a detailed discussion of the time and space wind data available for the Marshall Islands around the time of the Bravo shot, here.
AFSWP 895: fallout winds analysis by RAND Corporation for the 15 megatons Castle Bravo test on 1 March 1954 (the USS Curtiss was used as a weather observation ship which sent up balloons, which were tracked by radar to determine the wind pattern as function of altitude over the test site).  (Note that the Figure 6 caption is for Fig 7, shown below, and vice-versa!)

AFSWP 895 fallout in mushroom cloud of the 15 megatons Castle Bravo test on 1 March 1954 as determined by a RAND Corporation analysis. (Note that the Figure 7 caption is for Fig 6, shown above, and vice-versa!)
AFSWP 895: IBM701 computer summation fallout prediction method for 15 megatons Castle Bravo test on 1 March 1954 as determined by a RAND Corporation analysis.  This was developed by Stanley Greenfield of RAND Corporation, who states on page 348: "The first problem that was tried on the machine [an IBM 701 computer] was the Castle-Bravo shot", using the shot time winds measured from the USS Curtiss, a ship near Bikini Atoll.  The predicted Bravo fallout pattern is shown below:
AFSWP 895: IBM701 computer prediction of fallout using shot time winds for 15 megatons Castle Bravo test on 1 March 1954 as determined by a RAND Corporation analysis.  Notice that the fallout is predicted to essentially miss Rongelap Atoll (which is located from roughly 100 nautical miles East to 115 miles ESE, from ground zero).  Hence, there really was a wind shift that contaminated the islanders on the south of Rongelap (and nearby Americans on Rongerik Atoll, just to the east of Rongelap).  Even if the IBM 701 had been available to predict the fallout from Bravo on 1 March 1954, it would not have predicted the danger unless supplemented with a modern weather prediction including the changing wind pattern in the 6-7 hours following the detonation!
AFSWP 895: IBM701 computer prediction of fallout over Bikini Atoll using shot time winds for 15 megatons Castle Bravo test on 1 March 1954 as determined by a RAND Corporation analysis, comparison of measurements to predictions!
AFSWP 895: IBM701 computer prediction of fallout doses from a 50 megaton nuclear test as determined by a RAND Corporation analysis.  Note that the 1500 R dose would be reduced to a survivable 37.5 R by a protection factor of 40, the minimal specification for fallout shelters.
AFSWP 895 IBM701 computer prediction of fallout doses from a 1 megaton nuclear test as determined by a RAND Corporation analysis.  This is using the same model which successfully explained Bravo, and shows that with simple fallout shelters, fallout can be survived.
AFSWP 895: example of tabulated outdoor fallout areas for dose rates and accumulated doses from yields of 1 to 50 megatons.  Many different fallout models were compared in AFSWP-895, differences being due to different weighting in the activity distribution in the cloud and as a function of particle size, which affected how much activity came under the influence of winds blowing in different directions at different altitudes.  However, fallout distributions in the clouds were measured in detail in 1956 Redwing tests (using rockets with radiation meters and radio telemetry of data, see weapon test report WT-1315) and detailed particle size distributions (see WT-1317 and USNRDL-TR-314), so such disagreements are now resolved and fallout is very predictable with modern data from the 1956 Redwing series as well as modern weather prediction computer programs that include jet stream trajectory forecasts.  (Naturally, the ground deposited spectrum of fallout particle sizes at any particular location is biased in favor of the particle sizes that have a falling speed which results in their landing at that location, so this data needs to be backtracked to the cloud from a large number of representative locations to see what the overall distribution of particles is initially in the cloud when the toroidal downdraft has stopped operating.  Cloud samples are also biased in the same kind of way, because the largest fallout particles fall out before a sampling aircraft can safely get near the cloud.  Dr Edward C. Freiling's 1970 book Radionuclides in the environment, contains many papers graphically demonstrating this with data from cloud samples for Pacific shots in Castle, Redwing, and various 1962 Nevada surface bursts, such as Johnie Boy and Small Boy.  There is plenty of data, and shots on the differing soil particle size distributions in Nevada and the Pacific all tend to give a similar particle size distribution, closely approximating an inverse fourth power of particle radius, above 1 micron.)

UPDATES: 30 January 2018

Martha DeMarre of the Nuclear Testing Archive has also kindly supplied a PDF of the 1957 RAND Fallout Symposium (which we've uploaded to internet archive here), which contains an application of Anderson's dynamic fallout model to the 1.2 kiloton Sugar nuclear test in Nevada, 1951, to explain particle size distributions by tracking particles from the crater to their maximum height and then fall (rather than the usual false assumption that fallout occurs from a stabilised cloud).  This is listed on the DOE Opennet site but no PDF was previously available.  It also contains Schuert's demonstration that the time and space variation of the downwind wind structure correctly predicts the 3.53 megaton Zuni fallout pattern of Redwing (which is the only one of his four shot analyses which cannot be adequately analysed using merely shot time winds near ground zero; the other shots more easily predicted being Tewa, Flathead and Navajo), and summaries of the fractionation data for I-131 and several other nuclides in the 5.01 megaton harbour type surface burst Tewa at Bikini Atoll in 1956 in table 2 of appendix B:

1957 RAND Fallout Symposium: 5.01 megaton Tewa fallout radionuclide fractionation (depletion factor for volatile precursor decay chains) versus particle size and type for close-in samples from Bikini Atoll.  Note that I-131 is less severely fractionated than Sr-89, that the larger the fallout particles, the greater the depletion, and that spherical shaped particles have more severe fractionation than angular particles.  This is also seen in Tables 2 and 4 of USNRDL-TR-386 (AD232901) for the "Whim" sample of Zuni fallout (the test is identified in WT-1317): melted (spherical or "altered") particles had only 0.018 of the Sr-89 of unfractionated fission products, whereas unaltered (angular) particles has 0.65 and so were almost unfractionated, so they must have picked up the activity which was left behind after the melted particles were formed.  (For general data on Tewa, see the preliminary report of the test linked here.)  Note that the cloud sample data on Redwing fractionation is summarised in WT-1325, table 3.11 on page 47: fractionation was severest for lower altitudes in the cloud, where larger particles resided.  For example, only 0.51 of the expected unfractionated abundance of Sr-90 was observed at 41,000 feet altitude in the Zuni cloud, compared to a factor of 2 (enrichment) at 55,000 feet in the same cloud.  In table 3.2 on page 42 of the same report, for the 1958 Hardtack tests, it is shown that 1.31 megaton land surface burst Koa deposited 98% of its refractory Mo-99 within 24 hours, contrasted to only 64% of its volatile decay chain for Cs-137.  In the 9 megaton Oak surface burst test (effectively a land surface burst since the 15 feet of water above the reef at ground zero was trivial compared to the fireball radius), the corresponding figures were 89% of Mo-99 and 49% of Cs-137 deposited in 24 hours.  Volatile nuclides are concentrated on small, slow-falling particles located high in the cloud.

1957 RAND Fallout Symposium: Edward A. Schuert's predictions of the fallout hotlines for the 3.53 megaton Zuni test using different assumptions (shot time winds near ground zero, the space and time variation of the winds in the downwind areas through which fallout actually descends, and even vertical motions), compared to the ocean measured fallout intensities extrapolated to a land surface at 1 hour after detonation.

1957 RAND Fallout Symposium: Anderson's U.S. Naval Radiological Defense Laboratory dynamic fallout model analysis of the largest fallout particles (almost 2 mm in diameter) motions in the 1.2 kiloton Sugar test (Nevada, 1951).  Note that contrary to simplistic fallout models which assume that all fallout begins from the stabilised cloud 5 minutes or so after burst, no 1.95 mm diameter fallout particles remained airborne after 3.5 minutes in this test.  Anderson's model starts with dust being raised by the afterwinds from the crater, rising while that updraft force exceeds gravitation, then falling.  In this way, particles of different sizes rise to different peak altitudes in the cloud (the heaviest remaining mostly in the cloud stem, and the smallest rising higher).  This model thus provides the airborne distribution of particle sizes versus altitudes, and predicts fallout arrival times.

1957 RAND Fallout Symposium: Anderson's comparison of predicted accurate fallout distribution (solid line) being deposited 10 minutes after the 1.2 kiloton sugar test, with the inaccurate model prediction based on the false assumption of fallout beginning for all particle at 5 minutes from uniform mushroom distribution (dashed line).  Anderson predicts a smaller average particle size.

Neutron bomb secrets on Opennet: while searching Opennet, I found something else that is vitally important, already available for download as a PDF.  It's Johndale C. Solem's great 1982 Secret Los Alamos report LA-9004 on the neutron bomb, The ultra-low yield antitank weapon, the teeny tiny tacnuke, complete with declassified markings showing it was "Nuclear Weapon Data Sigma 1: Critical Nuclear Weapon Design Information", in a limited edition of just 79 printed copies:


LA-9004 from 1982, secret (now declassified with deletions of design information) states in its abstract (page 3) that: "Estimates of collateral damage indicate that such a device could be used in close proximity to civilian populations with minimal hazard."

LA-9004 then describes the kiloton W79 neutron warhead (44 cm long, 200 lbs including firing system, capable of being fired 32 km from a 8" howitzer), and explains correctly that the whole point of such weapons is to deter the concentrated blitzkrieg assaults that started WWI in 1914 (the invasion of Belgium by concentrated force) and WWII in 1939 (the invasion of Poland by concentrated force).  The principle of concentration of force can be deterred with nuclear weapons, thus preventing the invasions that trigger wars.  By forcing enemies to disperse their forces, any attacks that are made can be dealt with using conventional weapons like handheld anti-tank rockets (no use against concentrated firepower, but useful against dispersed forces), preventing invasion and WWIII:

"Denying an aggressor force the use of massed formations of armor is the single most important aspect of the W79."

LA-9004 then goes on to suggest a lower yield version of the W79 for use against individual tanks, like the Kennedy era portable 0.02 kt W54 that could be fired by individual soldiers, air burst at 15 metres altitude to eliminate local fallout, blast and heat collateral damage.  Page 5:

"Tank crews within 25 m of the weapon would be immediately incapacitated.  Civilian populations 300 m from the point of detonation would be completely safe. ... Beyond 300 m, exposed personnel might be temporarily blinded from looking directly at the detonation, but would suffer no burns to exposed skin. ... The effect of blast on civilian structures near the battlefield would be trivial.  Three hundred metres from the point of detonation windows would rattle but not break. ... the fallout would be expected to be confined to the battlefield itself. ... The principal advantage of such a device in reducing collateral damage from local fallout is that it simply does not produce much in the way of fission fragments or activated weapon debris."

LA-9004 then points out, on pages 7-8, that such a defensive low yield weapon with no significant risk of collateral damage is of no significant use to terrorists, contrasted to easy-to-procure alternatives.

UPDATE (5 February 2018): origins of fallout decay data in Glasstone's Effects of Nuclear Weapons

Martha DeMarre of the Nuclear Testing Archive has kindly supplied a PDF of the US DOE Opennet document NV0060036, the 15 April 1960 draft revision of the fallout decay activity section in Glasstone book The Effects of Nuclear Weapons, which was done by T. G. Brough and Dr Carl F. Miller of the U.S. Naval Radiological Defense Laboratory, California.  We have placed this PDF on internet archive, here.  The reason for investigating this is that the fallout decay graphs and tables in the 1977 edition of Glasstone and Dolan are identical to those in the 1962/4 editions, which differ from the 1957 edition.  Therefore, the current version was developed between 1957-1962, and this chapter revised draft by Brough and Miller from 1960 was clearly influential.  However, it is clear that Glasstone performed extensive additional changes to the 1960 draft before it was published in 1962.

In paragraph 9.6 of the revision, Brough and Miller explain: "the maximum radiation intensity of fallout from megaton detonations occurs at 50 to 75 miles downwind from the explosion centre."

They state in that paragraph that 1 fission megaton of fallout spread uniformly over 10,000 square miles would produce 410 R/hr at 3 feet height, 10 R/hr being neutron induced activity, and then they clearly explain in paragraph that this dose rate in reality the "fractionation losses" (i.e. the observed depletion of volatile nuclides from local fallout): "reduce the above mentioned radiation level at 1 hour from a value of 410 to 162 roentgens per hour."  This is far more specific and quantitative than the vague, entirely qualitative discussion of fractionation that made it into the 1962-1977 book!

Above: Brough and Miller's calculated revision to Glasstone's 1957 Effects of Nuclear Weapons fallout decay rate and accumulated dose graphs, which differ from those actually published in the 1962-77 editions!  Paragraph 9.111 at page 21 of their draft chapter revision also explains clearly than Glasstone's final version, just how the gamma ray energy of fallout varies with time and with the 0.105 MeV low energy contribution to the gamma ray spectrum caused by the neutron induced Np-239 content which is inevitable in dirty weapons with U-238 jackets that capture neutrons, and are not solely fissioned by neutrons (a fact essential for understanding how much shielding is needed to protect yourself against it, bearing in mind that fallout protection factors are calculated using the standard pseudo assumption that the gamma rays are like those from cobalt-60, which emits 1.17 and 1.33 MeV high energy gammas, a mean of 1.25 MeV, way higher than fallout):

"... the weighted mean energy of the gamma rays is about 0.92 Mev/photon at 1 hour after fission.  The mean value decreases with time during the first and second day after fission, and remains between 0.5 and 0.6 Mev/photon up to about 3 weeks after fission ... If the mixture contained neutron induced activities, such as U-239 - Np-239 in large amounts, the mean energy at early times would be much lower."

They even gave a table (Table 9.111 in the draft) showing that the mean energy of fission product gamma rays is 0.61 MeV (less than half the 1.25 MeV Co-60 average) at 24 hours, and 0.52 MeV at 2 days after burst, and remains around 0.5 MeV for the rest of the standard 2 week civil defense fallout sheltering period!  This is without the reduction caused by the very low energy gamma rays from neutron induced Np-239 and U-237.

These facts, deleted from Glasstone's published final version, reflect WT-1317 coauthor Dr Terry Triffet's June 1959 round table conference testimony on page 205 of the US Congressional Hearings on the Biological and Environmental Effects of Nuclear War, where he explains that this low gamma ray energy in dirty weapons increases the protective factor of shelters far above that usually assumed!

Brough and Miller's draft revision states at page 27 that their decay rates assume 8 MeV neutron fission of U-238, giving at 1 hour after burst 3600 R/hr per fission kiloton yield deposited per square mile, which is reduced to 1480 R/hr by fractionation, to which Np-239 adds 144 R/hr (this is a small percentage contribution at 1 hour, but becomes a much bigger contribution at 96 hours after burst due to differing decay rates of fission products and Np-239 which has a half life of 56 hours).