3.127, since at later times the fact that the center of the radioactive cloud and the center of the radioactive water are Separated by a considerable diStance makes interpretation difficult. The most probable position of the radioactive cloud is indicated in these plots as a Shaded area: the points not included in this area are considered to be dose rates arising primarily from racioactive water. These plots are too scattered for any precise cloud shape to be established; how- ever, they do indicate that the radioactive cloud for Wahoo is apparently continuous through its center whereas the center of the Umbrella cloud was relatively free of radioactive aerosol. These generai distributions of airborne radioactive material have alreacy been suggested in Section 3.3.2, and the presentation in this section is Simply a means of Summarizing the dose rate information from all stations in a single plot. 3.3.4 Transport Phenomena. Analysis of transport phenomena has been performed onlyto the extent necessary to interpret the gamma dose rate records. The material presented in this Section is again based upon an analysis of the GITR records and surge boundarypositions de- rived from photography as described in Section 3.3.2. The limitations of this preliminary treatment have been explicitly stated in Section 3.3.2. In general, however, there is reasonable agreement between data derived from photographic poSitions of the surge (Section 3.3.2) and that determined by use of the hypothetical surge center H (Section 3.3.3, and this section). Although the individual numerical values can only be accepted within wide limits of error, the internal consistency of data from several duferent analytical treatments is considered indicative of the validity of the general interpretation. Time of arrival (TOA) (see Appendix F) of radioactive material at a given station is the first obvious parameter for inspection; however, TOA may be defined as (1) time to the first pro- nounced increase in dose rate, (2) time to first peak in dose rate, or (3) time to some specified point on the increasing Slope of the first major dose rate peak. The last definition of TOA is undoubtedly best, yet the selection of the specific point depends on the base surge model assumed. By use of the computed dose rate curves for the approach of various hypothetical cloud models (Figures 1.9 and 1.10, and Section A.3), an average TOA may be defined as 38 percent of peak dose rate for any of the 60° cloud models, as 54 percent of peak for the 90° models over 1,200 feet in thickness, or as 74 percent of peak for 90° cloud models in the neighborhood of 400 feet thick. Accordingly, TOA has been read from the standard GITR records for 38, 54, 74, and 100 percent of the first major peak in dose rate. These values are tabulated in Tavie 3.14. The average of the 38- and 100-percent values is frequently used for the radiologically determined TOA (rad-TOA). This value is given in Table 3.11; as also indicated by this table, the photographically determined TOA (photo-TOA) occurs after the rad-TOA for Wahoo, whereas the reverse is true for Umbrella. A comparison between various photographically and radiologically determined approach velocities (discussed later in this section) indicates some preference for the thicker cloud models and a more definite preference for ali 60° cloud models. Since only rough general agreementis shown, no positive selection of cloud models can be made; however, if the 60° cloud models are accepted, the 38-percent TOA should be correct for most downwind arrivals. thus defined is presented for each shot in Figures 3.128 and 3.129. A plot of TOA Since this definition does not Satisfy all observed phenomena, a similar plot of TOA defined as 100-percent peak dose rate is also presented for each shot in Figures 3.130 and 3.131. A comparison of these plots reveals that they are relatively insensitive to the definition of TOA. After the break, the slope of the best Straight line for Wahoo is 15 to 16 knots, which is the Same as the reported surface wind speed at shot time. For Umbrella, arrival times at the more distant stations show considerable scatter, which may be dueto the effect of the atoll reef (Section 3.3.2). Visual 228 Pages 229 through 231 deleted.