a. Iceberg Drift Prediction 

 Model 



In investigating tiie effect of 

 SLAR, it was first necessary to 

 assess the possible impact of tlie 

 IBERG drift and deterioration 

 model on the number of icebergs 

 estimated south of 48°N, with its 

 introduction to IIP operations in 

 1979. Any significant effect on 

 the number of icebergs estimated 

 south of 48°N would necessitate 

 treating the IBERG and pre- 

 IBERG data separately in the 

 SLAR analysis. The IBERG 

 impact was analyzed by 

 comparing the number of 

 icebergs south of 48°N with the 

 number of sightings south of 

 48°N. Table C-1 shows the 

 number of non-growler sightings 

 south of 48°N from the 

 International Ice Patrol historical 

 data file for 1 960 through 1 984. 

 Since the number of icebergs 

 estimated south of 48°N does 

 not include growlers, growler 

 sightings were removed from the 

 data. The ratio of non-growler 

 sightings south of 48°N to 

 icebergs estimated south of 48°N 

 has three groups of values with 

 significantly different ranges: prior 

 to 1965 (9.9to 16.4), 1965 

 through 1978 (0 to 39.8), and 

 after 1978 (90.5 to 800). The 

 change in this value in 1965 is not 

 explicable through any change in 

 methods mentioned in 

 International Ice Patrol records. 

 After 1979, the change in this 

 ratio was due to the introduction 

 of the iceberg drift prediction 

 model (IBERG) to International Ice 



Patrol operations. It is important 

 to note that the ratio increased 

 dramatically in 1979, and that in 

 1 980, the "sense" of the ratio 

 changed; that is, the number of 

 icebergs south of 48°N became 

 greater than the number of 

 sightings received. With the 

 introductbn of this model, the 

 number of icebergs south of 

 48°N became an estimate based 

 on both sighting reports received 

 south of 48°N and icebergs 

 drifted south of 48°N by the 

 model. The model also made it 

 possible to more accurately 

 determine if a report was a resight 

 of an iceberg or an original 

 sighting. Since some resighting 

 was done before IBERG, the net 

 effect of introducing the model 

 should be to increase the number 

 of icebergs south of 48°N, due to 

 the icebergs drifted across 48°N 

 by the model. Examination of 

 Table C-2 , however, shows no 

 significant change in the 

 relationship between icebergs 

 south of 48°N and the iceberg 

 sighting ratio, a measure of 

 iceberg season severity 

 discussed in Part b. Therefore, 

 IBERG and pre-IBERG data will be 

 combined in the following 

 analysis of the effect of SLAR on 

 IIP iceberg detection. 



b. Icebergs South of 48°N 



In this section we seek to 

 evaluate the impact of SLAR on 

 the 1 983 and 1 984 seasons by 

 establishing an alternate indicator 

 of the severity of an iceberg 

 season. 



Overthe years, ships 

 transiting the IIP region have 

 furnished regular sea surface 

 temperature (SST) reports and 

 sighting reports for any iceljergs 

 encountered. These ship reports 

 provide a sample of iceberg 

 population data which is 

 independent of IIP detection 

 techniques. Since the numljer of 

 iceberg reports received is 

 dependent on both iceberg 

 density in the shipping lanes and 

 the amount of maritime traffic, the 

 number of reports alone cannot 

 be used to indicate the severity of 

 an iceberg year. Ships making 

 SST reports, assumed to be the 

 nrwst consistent iceberg 

 reporters, make their SST reports 

 in numbers independent of 

 iceberg density and provide a 

 measure of the annual traffic. 

 Therefore, by dividing the 

 number of ship iceberg reports by 

 the number of ship sea surface 

 temperature reports, a term 

 representing iceberg density, 

 independent of traffic, is 

 obtained. (Regression analysis of 

 SST reports versus icebergs 

 estimated south of 480N for 1 970- 

 82 yield an F value of .004, cleariy 

 demonstrating independence, 

 assuming nomrial distribution.) 

 We will call this term the iceberg 

 sighting ratio. 



Although the iceberg sighting 

 ratio is independent of the 



70 



