Figure B-9. Observed vs remaining length predicted by International Ice Patrol's deterioration model. The observed 

 wave height and sea surface temperature (SST) were used as the model inputs. The tick marks on the observed length 

 indicate observed instances of iceberg rollover 



100- 



140-1 



120- 



c 

 <u 



_l 



Ol 



q5 

 « 

 ^ 80 



O) 



c 

 c 

 ro 

 E 

 a> 

 tr 



e 



E 



X 



ra 



60 



40 



20- 



— 

 July 1 



Predicted Remaining Length 



Actual Length 



Observed wave height 



22 July 



attitude to the wave field. The 

 iceberg we observed rolled at 

 least seven times in five days. 

 Each time it rolled, a different 

 section of the iceberg was 

 exposed to wave erosion, 

 therefore requiring their 

 integration scheme to start over. 

 Since iceberg rollover cannot be 

 modelled, calving will not be 

 included in our model until an 

 acceptable scheme is available. 

 This will allow the predictions from 

 our model to remain on the 

 conservative side. 



Radar ranges and bearings to the 

 iceberg and HORNBEAM'S 

 position as determined by 

 LORAN-C were recorded every 

 hour. The geographical position 



of the iceberg was then 

 determined. 



On1 7 July, TOD #2632 was 

 deployed within 400m of the 

 iceberg in position 48°37.4'N 

 46O06.rW. The TOD drifted to 

 the northeast at a speed slower 

 than the iceberg. After five days, 

 the TOD was located 1 6.5 nm 

 away and bearing 252°T from the 

 iceberg (Figure B-10). This is 

 approximately upwind (using the 

 average wind for the period) of 

 the iceberg, indicating the 

 difference between the iceberg 

 and the TOD may be due to the 

 different leeway of the iceberg 

 and the drift buoy. 



The initial sighting position of the 

 iceberg was entered into HP's 

 iceberg drift model and allowed to 

 drift until OOOOZ 22 July. FNOC 

 winds and unmodified historical 

 currents were used as the 

 environmental inputs to the 

 model. The maximum difference 

 between the actual and predicted 

 position of the iceberg was 7.2 

 nm and occurred 30 hours after 

 the sighting. The error after five 

 days was only 4.3 nm (Figure B- 

 11). This preliminary analysis of 

 the drift data is encouraging 

 because the accumulated error 

 was so small. Further analysis of 

 this and other iceberg drift data 

 still needs to be done. 



66 



