18.104.22.168 Evidence of Changes in Arctic Pack Ice Thickness from Submarine Sonar
Estimates of thickness change over limited regions are possible when submarine transects are repeated (e.g., Wadhams, 1992). The North Pole is a common waypoint in many submarine cruises and this allowed McLaren et al. (1994) to analyse data from 12 submarine cruises near the pole between 1958 and 1992. They found considerable interannual variability, but no significant trend. Shy and Walsh (1996) examined the same data in relation to ice drift and found that much of the thickness variability was due to the source location and path followed by the ice prior to arrival at the pole.
Rothrock et al. (1999) provided the first ‘basin-scale’ analysis and found that ice draft in the mid-1990s was less than that measured between 1958 and 1977 at every available location (including the North Pole). The change was least (–0.9 m) in the southern Canada Basin and greatest (–1.7 m) in the Eurasian Basin (with an estimated overall error of less than 0.3 m). The decline averaged about 42% of the average 1958 to 1977 thickness. Their study included very few data within the seasonal sea ice zone and none within 300 km of Canada or Greenland.
Subsequent studies indicate that the reduction in ice thickness was not gradual, but occurred abruptly before 1991. Winsor (2001) found no evidence of thinning along 150°W from six spring cruises during 1991 to 1996, but Tucker et al. (2001), using spring observations from 1976 to 1994 along the same meridian, noted a decrease in ice draft sometime between the mid-1980s and early 1990s, with little subsequent change. The observed change in mean draft resulted from a decrease in the fraction of thick ice (draft of more than 3.5 m) and an increase in the fraction of thin ice, which was probably due to reduced storage of multi-year ice in a smaller Beaufort Gyre and the export of ‘surplus’ via Fram Strait. Yu et al. (2004) presented evidence of a similar change in ice thickness over a wider area. However, ice thickness varies considerably from year to year at a given location and so the rather sparse temporal sampling provided by submarine data makes inferences regarding long-term change difficult.