graphy. Consequently, an average charge radii of 376 feet were used, which compares favorably with the average charge radii of 387 feet computed for the Ivy Mike surface- level data obtained with electronic gages. The pressure-distance curve for these equivalent TNT charge radii was then scaled vertically by the NOL method for comparison with measured data, using the observed ambient conditions at altitude. The uncertainty of the measured data was such that it was not possible to correlate the vertical peak overpressures with the theoretical curves derived from the surface-level peak overpres- sures in this manner. Consequently, it was not possible to determine the best method of making an altitude correction to account for blast propagation through a nonhomoge- neous atmosphere for high-yield bursts. Those pressure data measured along the surface, obtained «on Shota 1, 2, 4, and 6 by using smoke-rocket and direct shock photography, are plotted in Figure 2.7. Gage data from Jangle Surface and Ivy Mike have been included for comparison and correlation. The data were normalized by scaling to 1 kt at standard sea-level] conditions, so that the composite free-air data scaled to 1 and 2 kt could be shown. A comparison to the 1- or 2-kt free-air curve for bursts was not strictly valid, shots involved an assumption fiection factor was presented the purpose of determining a reflection factor for surface since the hydrodynamic determination of yield for these of the factor of two. (Discussion of the surface-burst rein Section 2.3.5.) Figure 2.8 shows scaled arrival-time data obtained by smoke-rocket and direct shock photography, with the 1- and 2-kt com- posite free-air curve. Scaled data for both pressure and arrival time appear selfconsistent, as well as comparing favorably with Jangie and Ivy gage data. It seems justified to conclude, then, that cube-root scaling of blast data from events in this yield range ts valid. Part of the objective of the direct shock photography was to observe the formation and growth of any precursor which might occur. At this tre there was some doubt that the precursor would form on a surface shot. Actually, no precursor as such was noted; however, anomalous wave forms were recorded by the pressure-time gages. Observa- tions made of the film exposed on Shots 4 and 5 disclosed a dense water cloud following immediately behind the shock front. This cloud implies water droplets contained in the shock front and may explain the anomaly. 2.4.2 Base Surge. Early planning provided for the determination of the characteris- tics of the base-surge phenomenon for each of the shots. It was hoped that from such a study, scaling laws could be formulated to predict base-surge effects of surface shots with yields different from those of Castle. The base surge becomes of military signifi- cance when it acts as a carrier of radioactive contamination to regiong beyond normal falloThe extent to which this could occur from surface bursts, as well as the general dynamics of the phenomenon and the determination of scaling lawa, were the objectives of this study. The experiment was almost entirely unsuccessful, since the primary analytical tool, Photography, was rendered useless when it was decided to schedule the shots before sunrise. A minimum photographic effort was maintained throughout the series, from which it was determined that a base surge probably did form on Shots 1 and 2. This limited material prevented any detailed study anticipated in the early objectives. 2.5 CLOSE-IN GROUND ACCELERATIONS Study of ground motion produced by multimegaton devices detonated on the ground sur- face was planned for Castle to extend and supplementthose data obtained from Ivy Mike. 31