Long Term Activity Estimates for the Northern Marshall Islands

Page 1

This paper provides preliminary upper-bound estimates ofthe residual gama activity on the northern Marshall Islands due toUS. atmspheric testing at Bikini. These estimates are intended t o be ind;cative of the activity to be determined by up-corning deteiled surveys, Estimates are also provided for islands i n the Enewetgk atoll and compared uith the 1972 survey, files and fallout patterns is presented which serves t o delineate those northern Harshall islands which were uncontaminated by fa1 lout fro; Finally, an analysis o f wind prothe Bikini tests. After 20 years or so, the principal fission products of interest are SrgD and Cs13! whose characteristics are sumarized below. 1s0:ope Curieslkt of Fraction of H a l f Life Decay Mode Fission r t H+1 Total Curies . The fractional contribution of Cs137 tothe one-hour dose rate cSl3 ’ that is 0.41 times that for the inventory taken asa whole. some time after burst, when Cr137 isthe only remaining fission ptoduct where T is i n years. Note that beta activity is not being considered here on the presumption that the survey techniques distinguish between

Page 2

beta and gama. The above equation permits estimating the long term qamr,aactivity, provided there are one-hour dose rate measurements a t the locations of interest.4 The first step i n the analysis was t o compare the dose-rhtc estimates developed as prescribed above with recent surveys performed for the Enewetak atoll. This comparison would indicate the magn S tude of the difference due t o neglecting the migration o f the isotopes into the soil and plant uptake. Figure 1 4s a map ofthe Enewetak atoll showing the location of 3 islands chosen for the cofTpari&n--Alice, Janet, and Yvonne. Table 1lists the measured dose rate from the 1951-58 operations f o r these three islands as well as t h e 1972 estimates for The 1972 survey (reported i n N\‘00-140) provides average exposur: rates separately for &I3’ and Co6’. (This latter isotope i s not a fission product but results from weapon debris activation). soii depth for Alice and Janet. ‘ Itis important t o note that there evidently have been no cleanup activities(which would invalidate the comparisons discussed here) on A1 ice and Janet. Yvonne i s a different situation because o f construction and earth moving activities during the testing period. Large variations In exposure r a t e s occur on Yvonn?; thus, mean levels are misleading. .For t h i s reason, Yvonne will be dropped from the comparison. .. The dose rates can be comprred directly with the estimates o f fable 1. & expected, the estimates are high since among other reasons i t was assumed that the activity was all on the surface. The soil profiles of activity concentration versus depth can be used t o develop a pseudo dose rate by relocating the activity back tothe surface. of this value ’ with the estimate i s useful in that the difference i s hble2 provide; the CsI3’ survey data for Alice and Jan’et.

Page 4

Table 1. .Dose Rate Estimates for Enewetak ALS CE JANET YV0F;:iE I

Page 5

Selected Data from 1972 Enewetak Survey as a Function of Soil Depth~~

Page 6

then attributable not t o soil migration but rather to plant up tak c and other losses. To develop this pseudo dose rate,the following equation where a is the activity density inpCi / g,zis the depth i n cn, pis the soil density (1.8g/Cm ) and the factor of provides the cont o m-*. The dose rate for Cs : Table 3sum rize s the comparison between the estimated and measured : Csr3' dose r a t e and the..pseudo dose r a t e as well. As can be seen, the i estimate i s a factor o f about 20 higher than the measured value and ithat roughly half of this difference can be accounted for by nechznism other thzn roil migrction. estimates can be used t o provide bounding upper limits and that it mzgnitode by correcting fcr soil migration. The conditions for this This compErison indicates that simple . might be possible t o refine these estimates to within an order of a.) that for the location of interest, there had been no cleanup or major earth moving prior to the survey a n d b.)that the soil profiles would be similar to that found on undi sturbed Enewetak islands rtceivi ng fallout (such as Fig. 1409 o f "Sumnary o f Findings" Having compared dose rate estimates with survey results for ' hewetak, we can now turn to those islands i n the northern Marshalls that were contaminated by fallout from shots at Bikini. Because t h e estimating scheme being used requires the one-hour - dose rate as input, it i s $mportant t o first establish that off - site UIeasumnents were ma& in 811 Cases here there was fallout on the . $stands of interest. If these data are incomplete, estimations Cannot

Page 7

EST I MAT E DIRECT MEASUREMENT SOIL P”nOF1LE’ ISLAND - DI RECT MEASUREMENT INFERRED MEASURU.iEtiT*

Page 8

the Bikini tests. The last column i n most cases indicates that the wind directions precluded fallout on the islands. The det in it e exceptions are Bravo and Yankee. were in fact made. Bikini or other islands t o the east, so the test operations i n Table 1 cart be ignored. Table 4 summarizes the fallout pattern characteristics fro:, For Bravo and Yankee, off - site measurements tione cf the Enewetak shots resulted in fallout on :Figure 2 shows the Marshall Islands relative to the test loc a - tions. ?he Bravo fallout pattern has been reconstructed independently by AFSWP, NRDL and RAND using some modelling,while the Yankee pattern 3s based on extensive surveys. The one-hour dose rates f o r affected islands are given in Table5. All of the listed islands are outside the lowest dose-rate '(100R/HR) contour for Yankee (Rongelap is just barely); the levels ere stated only to' t h e nearest decade since extrapolation had t o be used. The range of values for Rongelap and Rongerik i s due t c the variation o f the Bravo pattern across the By and large, Bravo i s the predominant contribwtor. Table 6 provides 1977 estimates ofthe CsI3' dose rate for these islands. On the basis of the limited comparison performed for the Enewetak case, these values could be reduced by a factor of about The final part o f this paper 1s devoted t o identifying with high confidence which islands did not receive fallout from the Bikini ksts. Table 4 , as discussed above, indicates that only Bravo and Yankee definitely resulted in fallout on the Islands; this i s based on the use ofoff - site measurements t o reconstruct their respective fallout patterns. The other shots i n the Castle operation, f o r which there uere no off - site lneasurementr, apparently uere not a problem. flowever, a detailed investigation i s warranted and is reported on i n appendix. Also contained there i s an extrapolation of Yankee patterns .,to' a level consistent with background.

Page 10

To on gi Figure 2. Marshall Islands Affected by Bravo and Yankee Fallout

Page 11

Ai 1 i ngi nae Table 5 . One Hour Dose Rates for Bravo and Yankee Table 6 . CsI3’ Dose Rate Estimates for 1977

Page 12

On the basis of this investigation, the following islands are . extremely unlikely t o have received fallout from the Bikini or Enewetak tests at levels higher than the background exposure o f 200 &em/year: Kwa j ale in and any other islands circumscribed by the above. The following islands may have received some fall but from nuclear tests.It is unlikely that the intensities would have resulted i n an exposure o f nore than2 rem the first year; subsequent annual exposures would have been less than background: The following islands did receive fallout with intensities order o f decreasing residual activity : Taongi (bcsed on cloud drift only - no survey data available) the ib ove estimates, even when corrected for soil rnigraLion, can only &e considered pteliminary; they are very likely to be upper bounds. Note t h a t only C5I3' has been considered. The addition Of vation) ate necessary i n completing the estimates of the total activity

Page 13

The distribution of the activity in the s o i l , plants and organisms will not be determined by a simple survey of surface contamination. The estimates in this paper, along with such a survey, would be useful 4n determining such a distribution from the following kinds of additional a , ) water table height and variation b.) physical characteristics ofthe soil strata c . ) plant categories and root depth.

Page 14

PATTERNS FOR BIKINI TESTS ASSESSMENT OF WIND PROTILES AllD FALLOUT The Bravo a n d Yankee shots, as previously discussed, both deposited fallout on the islands east - of Bikini. reported contour level was not low enough t o circumscribe the total contour; this level was chosen because ft results i n an exposure t h e first year of about 200 mrem,which is about the annual background dose. Shown i n Figure 2is the southern periphery of the Bravo and Yankee In both cases, the lowest patterns relative to the location of the islands. b The other Castle shots are Romeo, Koon and Union ; off - site fallout measurements are not available so that their respective wind profiles have t o be examined. The Romeo winds a t H+3 and H+9 (DASA 1251) were not measured above 67,000 ft. Below this altitude the dominant direction of the profile is t o the north; while not.measured for the test, .the higher altitude ninds are uniformly tothe west. Thus i t is safe to state that the Romeo fallout did not reach any of the off-site Marshall Islands. Shot Koon winds were documented for a11 levels of interest. Except for near-surface, no winds had a northerly component t h a t would have carried any fallout to the south and east. I t can be stated with high confidence that Koon fallout carried t o the north and east, and did not reach any of the Marshall Islands. Shot Union presented a rather unique wind problem. Although . the lower altitude winds were from the east, strong northerly aid westerly components existed from 12,000 t o 50,000 feet. The influence of t h e winds i s not readily apparent without further examination. Therefore a crude reconstruction o f the fallout pattern was performed by determining the displacement o f 50, 100 and 20b particles which are initially assmed t o be a t cloud top and at cloud bottom. This permits the construction of an envelope ofall such particles i n the

Page 15

cloud. The H + 6 wind pro ile was used and constant f a l l rates of , 1 5 , ' .57 and 2.1 m/sec,respectively,nere used for the three particle s i t e s . (Including the altitude dependence of fall rate i s probablyan ovcrspecification, considering the uncertainty in the spatial variatior, of the wind). Shown 5n Figure 3is this envelope. affected by the Union falloufbut the other islands are outside the

Page 17

Long Term Activity Es tim tes For The Northern Ksrshsll Is lends This paper provides preliminary upper-bound estimates of t h e residual gama activity on the northern harshall Islands due toU.S. atmspheric testing a t Bikini. ind:cative of the activity to be determined by up-corning detziled These estimates are intended t o be surveys. Estimates are also provided for islands inthe Enewetak atoll and compared with the 1972 survey. Finally, a n anzlysis o f wind pro - - files and fallout patterns is presented which serves to delineate those northern Harshall islands which were uncontaminated by fallout frothe Bikini tests. After 20 years or so,the principal fission prolucts o f interest are SrgOa n d Cs13! whose characteristics are sumrtarized beloc:. Is3tspo Curie;/kt of Fraction of Half Life Decay Mode The fractional contribution o f 1 3 ~ ’ ~ ’ t o t h e one-hour dose rate 4s not the sBme as the fraction of total Curies a t one hour since the some time after burst, when Cs137 i s the only remaining fission product where Tisin years. Note that beta activity is not being considered here on the presumption t h a t the suwey techniques distinguish between

Page 18

beta and gamna. The above equation permits estimating the long term garma activity,provided there are one-hour dose rate measurements a t the locations of interest.4 The first step in the analysis was t o compare the dose-rktc estimates developed as prescribed above w i t h recent surveys performed for the Enewetak atoll. This compatison would indicate the magnitude ofthe difference due t o neglecting the migration of the isotopes into the soil and plant uptake. Figure 1isa map o f the Enewetak atoll showing the location of 3 islands chosen for t h e cornparison--Alice, Janet, and Yvonne. Table 1lists the measured dose rate from the 1951-58 operations f o r these three islands as well as the 1972 estimates f o r The 1972 survey (reported i n N\‘00-14’3) provides average exposure rates separately for and Co6’. (This latter isotope is nota fission product but results from weapon debris activation). soii,depth for Alice and Janet. ’ Itis important t o note that there evidently have been no cleanup activities (which would invalidate the comparisons discussed here) on Alice and Janet. Yvonne i s a different situation because o f construction and earth moving activities during the testing period. Large variations i n exposure rates occur on Yvonne; 1 ithus, mean levels are misleading. .for this reason, Yvonne will be dropped from the comparison. I fable 2 provide; the CsI3’ survey d a t a for Alice and Jaieet. The dose rates can b e compared directly with the estimates o f Table 1. AS expected, the estimates are high since among other reasons itwas assumed that the activity was all on the surface. The soil profiles Of activity Concentration versus depth can be used t o develop a pseudo dose rate by relocating the activity back to the surface. A conparison Of this value’uith the estimate i s useful in that the difference i S

Page 22

then attributable not t o soil migration but rather to plant uptake and other losses. To develop this pseudo dose rate,the following equation where a is the activity density in pCi/g, tis the depth i n cm, p is Table 3sum rites the comparison between the estimated and measured estimate is a factor of about 20 higher than the measured value and that roughly half of this difference can be accounted for by nechanisns other th zn soil migrztion. This compzrison indicates that simple estimtes can be used t o provide bounding upper limits and that it might …