Method 



The IIP iceberg sighting data base was 

 used as the source of data for this analysis. 

 This record contains all of the iceberg sightings 

 which are entered into the drift model during 

 the year. [This file is available from the Na- 

 tional Snow and Ice Data Center, Boulder, 

 Colorado.] It should be noted that this listing is 

 a subset of all the sightings received by IIP 

 during the year- the decision on which sightings 

 are entered into the data base is made by the 

 watch officer based on the iceberg density in 

 the area, the quality of the reported informa- 

 tion, and the geographic area. For example, 

 icebergs reported close to the Newfoundland 

 shore north of 49N have little probability of 

 drifting into the North Atlantic shipping lanes, 

 and may therefore not be entered into the 

 iceberg data base. 



The iceberg listing identifies each ice- 

 berg by a number, and gives the sighting 

 position, source, description, and last pre- 

 dicted drift position for the target. Resights of 

 the iceberg results in a new entry using that 

 same iceberg number. 



The iceberg listing was processed to 

 bin all sightings by latitude/longitude block. 

 The distance of the position correction was 

 computed, and averages computed for all 

 blockswheremorethanoneresightwasfound. 



Results 



The effect of using the DMPS system is 

 most clearly shown in the frequency with which 

 icebergs were resighted (Table I). In 1992, a 

 given iceberg had a 30.2% chance of an 

 adjustment to its position occuring over its 



Table I 



Iceberg Resights - 1992 And 1993 



Percentages are calculated based on the total number of 



icebergs identified for the given year. 



88 



