Figure D-5. Iceberg and TOD trajectories for Case III 



Results 



Figures D-9 through D-12 

 show the magnitude of the drift 

 errors as a function of elapsed 

 time for each of the four cases. 

 The II P error estimate of 1 0nm for 

 the first 24-hour period and an 

 addit'ional 5nm for each additional 

 24-hours of drift, up to a maximum 

 error of 30nm, is also plotted. 



In Case I (Figure D-9), the 

 system inputs result in drift errors 

 that increase rapidly and 

 persistently; after approximately 

 2.5 days they exceed 40nm 

 (~75km). The magnitude of this 

 error is 52% of the total predicted 

 drift. When observed currents 

 drive the model, the errors are 

 substantially reduced so that they 

 are nearly consistent with the 

 currently-used IIP error estimate. 

 In Case I, both the iceberg and 

 the buoy were in the southward- 

 flowing Labrador Current with 

 typical current speeds of 0.4-0.5 

 m/s. 



In Case II (Figure D-10), the 

 errors for the system/system run 

 were less than 1 2nm (~22km) or 

 22% of the total predicted drift for 

 the entire 1 04-hour drift period, 

 well below the IIP en'or estimate. 

 Using observed current and wind 

 data improves the results; after 

 1 04 hours the error is 2.5nm 

 (-4.6 km). The drift test was 

 conducted north of Flemish Cap 

 with typical current speeds of 0.2 

 m/s, approximately half that 

 observed in Case I. 



46°50' 



46''I0' 



1 20/OOz 



VOn 



Mno/ooz 



'^ .12Q'00Z 



AA9/00Z 



^^ 



s(oo* 



19/OC 



119/002#-Ji 



ns/ooz 



ICEBERG (▲ ▲) 



TOD38^m ( • • ) 



TOD 58m( O---0) 



46^0' 



46* 



73 



