Two TOD'S provided the 

 current data. They were 

 deployed, one drogued at 38m 

 and the other at 58m, 300m from 

 the iceberg on 27 April (117). 

 Approximately halfway through 

 the drift period, Ixjth buoys were 

 retrieved and redeployed close to 

 the iceberg to minimize the 

 separation between the iceberg 

 and the TOD'S. Upon 

 redeployment, the drogue at 38m 

 was set to 8m. The maximum 

 separation between the iceberg 

 and the TOD's, which occun-ed 

 during the first part of the drift 

 period, was approximately 35km. 



Case IV 



A 4-day drift [1-5 May 

 1985(121-125)]ofasmall 

 (60mx40mx10m) drydock iceberg 

 provides the data for Case IV. As 

 in Case III, the area of study was 

 south of Flemish Pass, and 

 EVERGREEN tracked the target 

 (Figure D-7) and obtained the 

 wind data (Figure D-8). Two 

 TOD's, one drogued at 8m and 

 the other at 58m, were deployed 

 onl May (121); they were 

 retrieved and redeployed at the 

 iceberg on 4 May (124). The 

 maximum separation tietween the 

 iceberg and the TOD's was 

 approximately 30km. On the last 

 day of the experiment, there was 

 a major calving event that left two 

 small icebergs. At this time the 

 parent (larger) iceberg had a 

 maximum wateriine length of 

 37m. 



Table D-1. Model Test Runs Summary 



Test Runs 



Table D-1 summarizes the 

 runs made during the model 

 tests. For each case, the first run 

 used the mean surface 

 geostrophic current field from the 

 IIP data base and wind data from 

 FNOC. This set of inputs is 

 referred to as system currents 

 and system winds. The remaining 

 mns for each case differed from 

 the first run only in that available 

 on-scene environmental data 

 (observed) were used to drive the 

 nnodel. 



The observed currents 

 were obtained from the TOD 

 trajectories by lineariy 

 interpolating to positions at 

 OOOOZ and 1 200Z each day, and 

 then calculating the 1 2-hour 

 averaged current. When wind 

 data were available, 12-hour 

 averages were computed for use 

 in the model. When no observed 

 wind data were available, FNOC 

 data were employed. 



For each run, the model 

 computed a predicted iceberg 

 position at OOOOZ and 1 200Z on 

 each date. The range and 

 bearing from the actual to the 

 predicted iceberg position were 

 computed for these times. 



72 



