differences. The special case of the bow GITR is discussed in Section 3.4.3. The prolon, sloping plateau extending from about 11 to 17 minutes 1S again most probably due to super ture turbulence, although a long record due to a baSe Surgetail is also possible (compare Station D 8.0). The earliest arrival of waterborne sources is about 12 minutes. If such sc did reach the ship, their arrival was masked by the last stages of base surge transit. La records (Figures 3.197 and 3.198) show the arrival of white water at about 1 hour. The b« and the port GITR register similar peak dose rates, but there is considerable difference | tween these records and that of the starboard GITR. Such differences could be caused by nonuniform distribution of radioactivity within the white water boundary. During Umbrella, the EC-2 was positioned at 1,650 feet crosswind with the port side fa surface zero. No ship movement was observed after Umbrella. The records (Figures 3. and 3.188} represent a central transit, which is close to being an inner edge transit. The | tral decrement appears to have been reccrded, but the high minimum doSe ratein the firs valley suggests contributions from 2 nearby inner base surge edge, although superstructu turbulence could also be a contributing factor. The prolonged record from 3 to 5 minutes most probably due to ship retardation. At 3 minutes, the white water boundary is approxi 1,000 feet radially beyond the EC-2; however, the record showslittle contribution from th source, AS suggested previously, the shielded location of the shipboard detectors probab. accounts for the reduced influence of waterborne sources. The records for the DD-474 during Umbrella (Figures 3.189 and 3.190) represent a cen cross section of the base surge, but possible effects due to the failure of the stern anchor Swinging of the ship must be considered. The DD-474 started with the stern toward surfa: zero and the port side facing the hot line; 20 minutes later it was 500 feet farther downwin with the bow toward surface zero and the starboard side facing the hot line. Very little c! in the ship’s position could have taken place during the transit of the downwind surge, but. lower dose rates recorded by the port GITR during upwind surge transit suggest partial si structure shielding. The slightly higher dose rates recorded by the bow GITR are also th result of differences in shielding. The prolonged gamma record from about 4 to 8 minute probably due to temporary retention of the surge by superstructure turbulence. A central decrement is, nevertheless, clearly recorded. The longer significant record of the starb GITR in comparison to the port GITR also supports the turbulent retention hypothesis, but difference is not consistently borne out by records from the other ships. Arrival of radio foam would be masked by base surge transit; however, its final departure is calculated at minutes. The rise in doSe rate between 22 and 26 minutes may, however, be due to such s At approximately 0.9 and again at 3.9 hours (Figures 3.201 and 3.202), there are relativel: den changes in the dose rate from waterborne sources suggesting rates of travel of 1 and! knots, respectively. The records for the DD-592 during Umbrella (Figures 3.191 and 3.192) also represent; central transit, which should be similar, since the DD-474 and DD-592 are at nearly the ¢ downwind distance 20 minutes after the shot. The DD-592 was oriented broadside with the board side facing surface zero. Because of the ship’s greater distance from surface zero times and heights of peaks are later and lower. The bow and platform GITR records trac! other closely as might be expected, Since they both represent relatively exposed instrume cations. No large increase in dose rate with increasing detector height is shown, indicati that both the bow and the platform instruments must subtend about the same solid angle of surge (Section 3.4.2). The port and starboard GITR records are lower than the bow and p. form records because of their relatively shielded locations. Again, the upwind transit las longer than would be expected on the basis of cloud photography. The prolonged recordis probably due to ship retardation although the ship’s crosswind attitude should minimize su retardation effects. The DD-592 was in white water from very early time. Later record: (Figures 3.203 and 3.204) indicate waterborne sources until 4.7 hours; thus, if the 6-minu' upwind white water boundary is used, a speed of 0.2 knot again results. 283