On the Estimation of the Distribution of Massive Marine Ice Features Derived from Moored Upward Looking Sonar Instruments

Using year-long, continuous measurements from modern upward looking sonar (ULS) data sets, potentially hazardous marine ice features can be identified and compiled as to their statistical distribution and frequency of occurrence. Previously, the focus of the analysis of these moored sonar data sets has been on identification of potential marine ice hazards in the form of the deepest ice keels, which extend to 20 m or greater along with occurrences of rubbled/hummocky ice features with horizontal scales of up to a few to several hundred metres. A more robust, geometrical characterization of potentially hazardous marine ice features is presented in this paper to identify episodes of continuous ice draft measurements exceeding a threshold value, for which the cross-sectional area of the ice draft and horizontal ice distance is computed, using a user-selectable minimum threshold value (e.g. 1.2 or 2.0 m). Based on several hundred or thousands of individual features identified, the maximum and average values of the cross-sectional area are computed. Using an assumed statistical approach for the horizontal geometry of the individual ice features, the total volume of these large marine ice episodes, can be estimated as the product of the cross-sectional area and the single measured width, with over 1,000 features each year exceeding the median value of 82,000 metric tonnes for a 2.0 m minimum ice draft, being realized in the mooring data sets north of Fram Strait off northeast Greenland. The numbers of massive ice features exceeding nearly the same median value is much reduced at about 160 in the Canadian Beaufort Sea.