Statistical properties of Arctic sea ice morphology derived from sidescan sonar images

Using a transparent overlay technique, the authors have classified a 140km long stretch of sidescan sonar imagery of the Arctic sea ice underside (700m swath width) into regions occupied by unique ice-types. The categories used were multi-year ice, first year ice, pressure ridge, apparently open ('dark') lead, apparently refrozen ('bright') lead and randomly deformed ice. The technique was validated by comparing these classifications with data obtained simultaneously using a narrow beam upward-looking sonar. The pattern of ice-types thus observed was tested statistically for associations between adjacent regions, the distribution of lead widths, spacings and orientations and the distribution of ridge spacings and orientations. Significantly non-random sequences of ice-types were observed, lead orientations and spacings were clustered and were set to preferred values and ridge spacings fitted a negative exponential distribution. The extent of uniformly deformed ice (with no apparent linearity) was much greater on the underside of the ice than has ever been observed on the upper surface.