Table 4 

 New IIP Iceberg Population Severity Classes 



Light 



Moderate 



Extreme 



x; 



<300 icebergs south of 48°N 

 300-600 icebergs south of 48°N 

 >600 icebergs south of 48°N 



^ 



plot, Marko, et al. (1994) suggested three 

 iceberg severity classes (See Table 3). Their 

 plot is updated in Figure 3 using revised sea 

 ice dataV 



The good correlation coefficients be- 

 tween sea ice and iceberg populations (Figure 

 3) highly suggest that, with few exceptions, 

 the iceberg counts since 1 963 are reasonable. 

 The one notable exception is 1984, a year 

 suspected to be greatly overestimated be- 

 cause of Ice Patrol's inexperience with a new 

 sensor (Thayer (1984)). The good correla- 

 tions suggest that the quality of the data is 

 relatively consistent between the two eras. 

 This is another reason not to make conclu- 

 sions on an era average. 



The bilinear response in the iceberg- 

 sea ice scatter plot and the good correlation 

 validates the Marko, et al. (1994) iceberg 

 severity classes. Their definitions are very 

 close to the Alfultis (1987) definitions except 

 for the greater-than-900 class. There is no 

 evidence of a fourth category in this correla- 

 tion. The populatiorLsevarityLdefinitions-sug- 

 gested by Figure 3 are contained in Table 4. 



Table 5 

 IIP Ice Season Length 

 Severity Classes 



Short 



Average 



Long 



< 105 days 

 105-1 80 days 

 > 180 days 



^ 



^ 



The slight disagreement between the 

 Table 3 and 4 lower thresholds is due to 

 slightly different sea-ice data. Figure 3 sug- 

 gests that the lower (light) threshold is be- 

 tween 200 and 500. A threshold of 300 was 

 chosen for consistency with Alfultis (1988). 

 The purposeful use of term "moderate" was to 

 avoid any connection to the terms "mean" or 

 "average". 



ICEBERG SEASON LENGTH SEVERITY 



A severity index that has not been pre- 

 viously investigated is the length of the ice 

 season. This is roughly the amount of time that 

 -icebergs are present south of 48°N. The 

 season length correlates well (r=0.7) with sea- 



' Marko's (1994) correlation looks slightly different due to a different sea-ice data. Marko 

 used tha ice extent in three different bands 45°-47°N, 49°-51°N and 53°-55°N as estimated 

 from Prisenberg and Peterson (1990). I used the entire ice extent from 45° to 55°N. 



51 



