above the tropopause increased in speed to the highest altitude of the available wind information for the Bravo shot, while for Mike the easterly winds decreased in speed and ultimately changed to westerly winds. The easterly winds in the trade-wind layer, the moist maritime air mass lying near the sea, extended up to about 20,000 feet during the detona- tion of the Mike device, while for the Bravo shot they were below 10,000 feet. Between the trade-wind layer and the tropopause, one normally finds westerly winds. During the Mike shot these westerlies were temporarily interrupted and became southerly winds, while for the Bravo shot they were toward a more normal bearing. In Fig. 2. is found the approximate area covered during the early days by that part of the nuclear cloud from the Mike shot which was located below the tropopause. The shaded areas in Fig, 2 have been deduced from meteorological considerations alone, and, in manycases, are subject to considerable uncertainty. Shading was discontinued when the meteorological data no longer warranted any reasonable estimate of the path. The light winds and sparsity of upper-wind observations have made tracing the upper tropospheric portion of the Mike cloud particularly uncertain. For this reason, the time of passage across the North American mainland is unknown. Tracing was discontinued on 7 November. The tradewind portion of the nuclear cloud appears to have split south of Japan, the upper portion (near 20,000 feet) curving around a Pacific high cell and entering the United States about 9 November. The estimated meteorological path of the Bravo cloud is shown in Fig. 3. The upper tropospheric portion of the nuclear cloud was traced to the Central Ameri- of the two nuclear tests, suggests a path that would travel around the carth at about the same latitude as the point of origin. It is interesting to note that in no case was it imperative to rely on stratospheric transport of the nuclear debris to account for the earliest arrival at any point, for the transport of the nuclear cloud in the troposphere appeared to ac- count for the first observations of radio- activity. An attempt to determine the earliest arrival time at the ground at each point of observation has been undertaken. The results, which are shown in Figs. 2 and 3 as the numberof days after the shot day, should in many cases be viewed with caution. First, in many of the stations in feature unusual for the region during any season. There has been no attemptto track the stratospheric portions of the atomic cloud because of the sparsity of wind observa- tions at these altitudes. Evidence from numerous isolated high-level winds, not necessarily obtained during the periods 14 SEPTEMBER 1956 The comparatively small values ob- mage Sn + 110,000 “+ + 100,000 Sy + + + + +> +o +s 70,000 t~, aa g 60,000 +n _, ~ ° £ 90,000 Fallout 80,000 It is noted that, in accordance with the meteorological estimates, the fallout over the United States progressed roughly from west to east during the Mike shot. Fallout from the Bravo event did not appear at the West Coast stations in the United States until 2 weeks after one of the cloud protuberances entered the cen- tral United States. Of perhaps greatest interest, although also of greatest doubt, in the Southern Hemisphere. Thus, for example, a literal interpretation of the chart reveals that every station in the Southern Hemisphere showed an earlier arrival time than did the United States West Coast stations for the Bravo case. of interest are the compara- tively late arrival times for the mid-Pacific stations west of the Hawaiian Islands during the Mike fallout. These stations were south of one branch of the nuclear cloud and north of the other. The actual fallout at each station and an analysis of the data are shown on Figs. 4 and5. The units are cumulative decayed beta activity for the first 35 days following each event and are approximately equivalent to millicuries per 100 square miles (the values have not been corrected for the efficiency of the gummed film.) Several features that differentiate the two maps should be noted. First, an aver- age value for all United States and Canadian stations was obtained for the Mike shot, as opposed to values for individual + ¢ rival time of the cloud at manystations coincided with rainfall, suggesting that the nuclear cloud may have been overhead some time earlier but that precipitation was required to bring its activity to earth. Also tion not far from the Marshall Islands whichis not typically present in March. The shallowness of the trade-wind layer during the Bravo shot is an example of a marily in an attempt to locate stations in rainy areas. In manycases, when the period of record is incomplete or the data are suspect, parentheses have been placed around the number. No attempt has been madeto reconstruct the isolines for the fallout that occurred within the first 24 hours of the shot. despite elaborate precautions,it is likely that some gummed films were contaminated during handling. Finally, as noted in the second paragraph the apparentar- are seasonal and in part due to the spein November the mid-tropospheric westerly winds are not as strong as they are in March, and they are located farther north, on the average. Further, in Novemberonefinds an anticyclonic circula- Finally, as is evident, the network was expanded between the two events, pri- activity was so low that it made the arrival date almost meaningless. Second, are the comparatively early arrival times cific metcorology for the shot days. Thus, tained from transport vessels for Bravo. the Southern Hemisphere, the deposited can area by about 5 March, and an offshoot extending northward into the United States at about 20,000 feet was detected approximately 1 week later. Differences between the paths of the Mike and Bravo clouds are evident from Figs. 2 and 3. In part, the differences stations during the Bravo shot. Second, the isolines located between points on the West Coast of the United States and points in the Western Pacific Ocean are also based on fallout observations ob- z 5 + LoS +. pe ty 50,000 Vt or 40,000 T ot toa J 30,000 ; ot 20,000 Ns web _K ae 10,000 ad KY 0 N + N “te wre MIKE BRAVO. Fig. 1. Upper winds at shot time. Arrows ! blow with the winds, and barbs indicate wind speed; full barb, 10 knots; one-half barb, 5 knots. 475