the Crocker estimate at about 10 hours post detonation and virtually identical activities were estimated after 100 hours post detonation. £. Exposure Rate from Surface Activity The mean exposure rate estimate from all the nuclides given in Table 5 and present on day 0.5 post detonation of the BRAVO device was 1.1 x 19! R nv, This estimate included the contribution from noble gas nuclides which would not have remained on the surface with fallout particles. Exclusion of the noble gas activity yields 9.7 R nw? for the exposure rate at day 0.5. Due to surface roughness effects the unweathered exposure rate estimate of 47 mR h 1 could have been an underestimate which would have resulted in an underestimate of surface activity on day 26 post detonation. Surface roughness effects were approximated by comparing Beck's values for mR nt uci i m” for a smooth flat plane, to a plane where activity was distributed depthwise with a relaxation length of 0.16 gm cm. This translates into an un- derestimate of the activity present by 15%. Based on Bikini Ash composition and negieccting the impact of surface roughness effects and unaccounted for nuclides, the estimate of average exposure rate at Rongelap Island on day 0.5 at one meter above the surface was taken to be 9.7 R nl, This was significantly different from the extrapoled value of 5.8 R nt on day 0.5 derived from Miller's decay estimates (see Table 3). The decay exponent derived from Bikini Ash data was -1.42 for the period 2.2 days to 26 days post detonation. The value for the decay exponent for the period 12 hours to 2.2 days was -1.23, from 9 hours to 12 hours it was -1.31, and from 5 hours to 9 hours it was -1.41. These early values for the decay expo- nent were significantly different from those derived from Miller's decay estimates (see Table 3) and indicated much more rapid decline in the exposure rate. 21