Castle-Bravo Air Concentration and Deposition Patterns from a 3-D Particle-in-Cell Code (Draft)

Page 1

The JIATFi E!V-ADPTC code suite has been extensively modified to give the total external dose from the detonation of the Castle-Bravo nuclear test at Bikini -Atoll until evacuation of the inhabitants of nearby atolls. The aavantcges of this code suite is that it uses all the observed winds (in a ,mass+onservstion sense) at and ~fter the detonation to provide dose rates and doses d{[e to passage of the debris cloud and to the time-integrated deposition up to evacuation time. Previous assesstmencs have given the fallout pattern (deposition only) at time H+l hours. The present code formulation gives excellent agreement with the estimated total external close (based on measurements) to people on Rongelap and .Ailingin~e

Page 2

The author is indebted to Rol[ Lange, Leonard Lawson, and Hoyt \’/alker of La.,,vrence Livermore National Laboratory for their assistance in developing tine suite of computer programs used in the calculations. Grateful thanks are also due of 13rookhaven National Laboratory rate data for the Castl-Bravo test. to Nathaniel Greenhouse* and Edward Lessard for supplying me hfith meteorological and dose “Present a.ft’ilia~ion: Lawrence Berkeley National Laboratory, Berkeley, California

Page 3

Operation Castle was an atmospheric nuclear test series conducted in ths Jlarsha]l Elands from Nlarch to NIay of 1954. The most notorious test of the series ,was Bravo, a 15 megaton(l] thermonuclear explosive. ‘The top of the resultant debris cloud reached to nearly 35 km at stabilization time. [11 Because of an unexpected shift in mid-tropospheric wind directions following detonation of Bravo, the fallout pattern, instead of heading in the predicted northeast direction, had an easterly alignment. As a result, persons on the atolls of Rongelap and Rongerik were exposed to relatively high levels of fallout from the nl[clear explosion. Prompt action was taken by U. S. Task Force personnel to evacuate the natives of these islands. Some of the natives on Rongelap, the closest to the detonation point, suffered temporary nausea and minor skin burns. None exhibited any medium or long term effects from their exposure... Ho’wever, after about 10 years, those Rongelap natives, who were young children in 1954 developed non-m aligment nodules on their thyroid glands. Since then the occurrence of similar nodules among the Utirik natives has been reported. The rate of occurrence has been higher than would be expected statistically. The purpose of this report is to calculate deposition and surface air concentration plots, using a thre+imensional particle-i tieil suite of codes to estimate the doses at the islands from which the natives were evacuated. \Ve will aIso consider the dose from rainout as part of the debris cloud crossed the atolls. Finally, the calculated time history of air concentrations on the downwind islands will be presented for several

Page 4

Some of the better known patterns apoesr in Ret. 1 and were prs?arsd b:~ (:i) the .Air Force Special Wespons Project, (b) the NavJl Radiological Defense Laboratory, and (c) the Rand Corporation. A comparison of these three patterns sho.vs significant differences in the maximum dose rates, as well as the shapes of the contours. This is due in large part to the subjectivity involved in the calculations. Portions of the Rongeri!<, and Utirik, as well as a crude estimate of the dose rate received by the Japanese fishing ship, the Lucky Dragon. The remainder of these patterns were obtained using the observed winds in a subjective manner to bend the pattern and achieve an approximate mass baiance. The Rand contours used estimated winds between Bikini and Rongelap. These winds were obtained from interpolation of streamline analyses at several levels at By contrast, the altered versions of the NIAT’HEW”-ADPIC codes used in this report allow us to use the observed winds at different Iocations and different times after detonation. No artificial bending of the pattern is required. The only subjectivity lies in the selection of code input parameters. At all times, the codes automatically assure conservation of mass. The suite of codes develoged for the Atmospheric Release .Advisory Capability (ARAC) were extensively modified in order to incorporate a larger number of upper air wind levels. All pr”ior uses of the codes have been to handle calculations for releases that did not rise higher than a few kilometers. Also, codes do not involve sophisticated gravitational fall velocity they include time-integrated deposition. the standard ARAC calculations, nor do

Page 5

from cell to cell. After modification, all observed upper air wind data from 10 m to oi]e to three observing stations for each time. The obvious advantages of (over most other fallout codes) is that mass is not permitted to accumulate in any of the cells and winas are available for each 3-D ceil intersection f’or four times. The :\iATHE\\ winds are used then by the modified ADPIC [3] particle-i n-c ell code to caIculate the transport, diffusion, and deposition of an instantaneous source. Modifications required of ADPiC, to handle the Bravo test, consisted of allowing more upper air input than is used in typical A.RAC assessments. Furthermore, since particles falling from the stratosphere undergo a large increase in air density, it was necessary to particles; this correction follows modifications were to incorporate add a turbulent wake correction to the larger the method set forth by Llci)onald; Other a tropical atmosphere into the fall velocity calculations and to make the particle activity increase as the cube of particle radius. Finally, time-integrated deposition was added; this allows calculation of the total dose from detonation ti,me to evacuation.

Page 6

Surface metsorologicd observations were avaii~dle from some atolls and from the U.S.S. Curtiss w~hich cruised south of aiki~i; hoivever, since the Iarger particles fall rapicly from the debris cioud to the surface and spend little time near the surface, not many surface reports were used. Of far greater importance are the upper air ~vind observations taken at four sites near Bikini atoll. Other si~~ii’icant input data consisted of a fiat topography, cell sizes of 34 km (east-west) by 17 km (qorth~outh}, and 1 km in the vertical, stem and cap debris cloud geometries at stabilization time, source rates for both gross fission products and selected individual nuc!ides, and particle size spectrum parameters. l%e tim~integrated external dose pattern (in rack) due to gross fission woducts from detonation time to evacuation time of Rogelap atoll (51 hours) are. . shown in Fig. 1. The numbers next to Ailinginae, Rongeri!{, and Utirik atolb are integrated values up to the time peooie were evacuated from those atolls. For comparison, the value of total dose, estimated by Dunning [5J and 5traus46] are given in Table 1. Note that the agreement is very good for Rongelap and Ailinginae atolls. However, calculations for Rongerik and Utiri!< nre :lt odds ‘with eariier estimates. The code calculations for Rongerik are higher, while those for Utirik are lower. This variation appears to be in part a problem of “tuning”; also a l)Ossib~e variation in ‘wind directions and speeds at late times when the only wind observations were from the U.S.S. Curtis, south of Bikini (some distance from the atolls of concern) may be an explanation.

Page 7

Tahie 1. Comparison of gros3 fi~~icn product total externai drojes for Evacuation Time Calculations Estimates \\’ith the modified MATHEIV-ADPIC code suite, it is possible to calculate the instantaneous immersion dose rate from gross fission products as a function of time. 17nis can be done for any time intevel. Fi=wres ?a to 2f show surface immersion dose rate cmtours for every three hours from one hour after Bravo cloud St~k)ihZiil,iOII time to Hi-16 hours. Note that after an easterly traverse, most of the cebris reaches the trade wind level; the contour pattern moves south and fina~y towara the southwest. Calculations were made of instantaneous and tim~inte~ated concentrations at 2 m aaove the surface for the several ato~s affected by Ca~tle-Bravo. The calculations agree lwell with observations at Rongelap and Ailinginae atolls, but are too high at Rongerik and too low at Utirik atoll. The surface concentrations for .+toll’ are presented in Figures 3a to 5e. The time of arrival of the first Bravo debris is in agr~ement with reports made by the inhabitants.

Page 8

cccurred for a short time after Bravo was detonated. The crew of the Japanese fishing ship, No. 5 Fukura Alaru (LucLv Dragon), while fishing do’wnwind just o~tsi~e the exc!usion zone, noted that the initial fallout on their ship was accompanied by “a ~7] it is unlike!y that this was a continuation of theIight rain or drizzle.’ self-induced rainout, some two or more hours after Bravo’s detonation; it was probably a natural rain system superimposed on the debris cloud. Another report of rain during Bravo fallout was made by a group of Rongelap natives after evacuation~g] They lived in Rongelap Village, on the southern part of Rongelap Atoll, and stated that it “rained a little” during the aiternoon of Llarch Another interview with an A,merican Air Force radio operator [8] who had ~een on Rongerik Atoll prior to evacuation disclosed that “rain commenced about 2100 [LST] and continued for 30 minutes.” Finally, the S. S. Roque, o’wned by Tlicronesian Lines, left Kwajalein at 0945 LST and arrived at Utirik at about noon on March 2, 1954. The ship left Utirik (apparently a few clays later) and arrive at Majuro Atoll on March ‘7. A radiological survey at hlajuro disciosed radiation readings of 10 to 30 rnr/h on March 7. The ship’s captain mentioned that he had encountered rain squal.is during his voyage, but was not specific about where or when. It appears certain that the S. S. Roque e ncountered Bravo fallout, possibly accompanied ~pproaching or whiie in har’bor at Utirik. If 10 mr/h c!ays earlier (when the Bravo debris cloud passed estimated at about 100 mr/h. by rain sho’wers, either while are “grown back” to five or six near Utirik), the dose rate is

Page 9

Extensive modification of the ~.IATHEi’i-.ADPIC code suite has pKK!I:ced contours of Castle !3ravo accumulated and time-integrated deposition for gross fission products. Through the use of dose conversion factors, these contours have been converted to dose rates and totai doses uo to the time of evacuation from the atolls affected by the debris cloud. In addition, both instantaneous and time-integrated surface concentrations have been calculated. For the nearest atolls, the calculations agree well with the measurements and total dose estimates based on these measurements. At the more distant atolls the agreement is not as good, indicating the need for more “tuning” of the code input parameters. The internal dose to the inhabitants of the affected ~tolls have not been made in t!~is report. Interviews with natives of Rongelap Village and Aiiinginae [8] indicate that many people ate fresh seafood and drank water from cisterns follo’.ving contamination of their islands. Although there is no direct evidence that those at Utiri!; ate and drank contaminated food and water, it seems li!<ely that they did since the dry deposition from Eravo was considerably less than at atolls to the west. However, the previous section indicated that rain probably occurred during the time o f fallout. This would result in wet deposition, producing local doses 10 to 50 times greater than in those areas where rain did not occur. This effect COUld have resulted in development of thyroid nodules in those L’tirik residents who consumed contaminated food and ‘.vater.

Page 10

Disgersal of Atmospheric! Pollutants and its Comparison to Reqional Tracer Three Incidents of Fallout,” in Proceedings of .4 Symposium on RaCioloqi~a~ Protection of the Public in .\ Nuclear ?,lass Disaster, Interlaken, S’.vitzer!artd, 6. Strauss, Lewis, Statement made at hearings of the Joint Committee on Atomic (. Lapp, Ralph E., The Voyage of the Luc!;v Dragon, Harper and Brothers, New 8. Sharp, Robert, “Exposure of llarshall Islanders and American Jlilitary Personnel to Fallout, ” \’iT-938, Operation Castle - Project 4.1 Addendum, Mookhaven National Laboratory

Page 11

1. castle Bravo time-inteqgated extennal q=~a dose from qross fission products. Contours zre for an H+51 hour evacuation time from Rongelap. Wmbers added for otier atolls areintegrated to the appra?riate evacuation time far those atolls..