survey observations. The positions of the concentrations are ad- 

 vanced to estimated positions for the date desired, based on average 

 drift due to current and wind, taking into account bottom topography 

 and sea ice conditions. The distribution of the concentrations, namely 

 their locations relative to the coast and current axis, is a very im- 

 portant consideration. Tliis, of course, must also be estimated based 

 on last oibservations and estimated drift trend due to wind, while tak- 

 ing bottom topography and the current system into account. If the 

 estimated along-current position of each half of the berg crop is incor- 

 rect by 100 or 200 miles this will not introduce any great error in 

 evaluating the effect of pressure distribution, as there will normally 

 be little or no significant difference in this relatively short distance. 

 However, if the estimated berg distribution is significantly inaccurate, 

 the effect of a monthly pressure pattern might be incorrectly classified. 

 For example, if most of the concentrations of bergs were closer ashore 

 than estimated, the effect of along-current winds might be unfavora'ble 

 due to bottom topography, whereas favoraJble classification were in- 

 correctly assigned. Likewise the reverse situation might occur when 

 winds were classified as unfavoraible due to an erroneous assumption 

 that bergs were close ashore whereas in reality most of the bergs were 

 well offshore and free to move. When no oibservations are available, 

 it must be assumed that the groups are centered near the axis of the 

 current unless previous pressure distribution patterns indicate other- 

 wise. The point is that the along-current locations of groups and 

 the locations of concentrations relative to the current axis must be 

 estimated as accurately as possible in order to correctly evaluate the 

 effect on their drift due to wind. 



The importance of the bottom topography of the continental shelf 

 and the coastal configuration from Baffin Island to the Grand Banks 

 of Newfoundland on the iceberg problem has already been mentioned. 

 Large bergs will normally ground in waters less than 90 fathoms, and 

 medium bergs will generally ground in waters less than 50 fathoms. 

 A study of the shelf waters from Davis Strait to the Grand Banks 

 reveals the existence of many shoal areas of 100 fathoms or less extend- 

 ing well offshore and well into the colder arctic portion of the Baffin- 

 land-Labrador Current system. There are many natural traps along 

 the berg route from Baffin Bay to the south, and some bergs are 

 permanently trapped and removed as a threat to shipping near the 

 Grand Banks. Perhaps the two most important permanent traps are 

 Hudson Strait and Belle Isle Strait, where during some years many 

 bergs are trapped. Aside from these two areas, larger bergs are 

 generally immune to permanent trapping as their deep draft normally 

 causes them to be grounded before getting into the many permanent 

 type traps including the numerous bays, harbors, and other indenta- 

 tions and channels along the irregular coastlines. The current axis, 



49 



