Automated Detection of Hazardous Sea Ice Features from Upward Looking Sonar Data

Abstract

Upward-looking sonar (ULS) instruments provide extended continuous measurements of ice thicknesses and ice velocities data that are important for establishing metocean design criteria related to oil and gas operations in areas with seasonal or year-round ice cover. This paper describes the development of algorithms for the detection and measurement of hazardous ice features including: large individual ice keels with thicknesses of 5 to well over 20 m; long sections of thick hummocky (rubble) sea ice; and occurrences of multi-year ice floes. Large individual ice keels are detected using an ice draft threshold technique to identify very thick ice floes which are then categorized as to total width using a Rayleigh criteria and/or a minimum user specified threshold value (e.g. 2 m). The detection of thick hummocky ice is based on minimum criteria of ice draft data segments having median values exceeding 2.5 m and segment lengths exceeding 100 m. For qualifying segments, a selection parameter γ, defined as the 90th percentile over the 50th percentile value of ice drafts divided by the standard deviation was computed; hummocky ice is characterized by γ > 2 and is also very common for 1.5< γ <2. Results from the ongoing algorithm development for detection of multi-year ice features will also be discussed. Ice velocities can also pose difficulties for offshore oil and gas operations in terms of floating drilling platform station keeping when particularly large ice speeds occur and/or ice drift directions changing rapidly or erratically.