Layer | Fill | Outline |
---|
Map layers
Theme | Visible | Selectable | Appearance | Zoom Range (now: 0) |
---|
Fill | Stroke |
---|---|
Collaborating Authors
Results
Parallel Channel Tests During Ice Management Operations in The Arctic Ocean
Lu, Wenjun (Norwegian University of Science and Technology) | Lubbad, Raed (Norwegian University of Science and Technology) | Løset, Sveinung (Norwegian University of Science and Technology) | Skjetne, Roger (Norwegian University of Science and Technology)
Abstract During ice management operations, creating narrow parallel channels with icebreakers can effectively reduce ice floe sizes for the protected vessel/structure. Yet too narrow channel spacing requirement shall lead to excessive or even unpractical ice management operations. Empirical experience shows an almost 1:1 relationship between ‘downstream floe size’ and ‘parallel channel spacing’ while designing an ice management operation. However, before this paper, there exists no dedicated parallel channel tests with strictly controlled channel spacing and sufficient instrumentation to quantify such relationship. In this paper, we report two parallel channel tests, which have been conducted in September 2015 during an expedition to the Arctic Ocean with icebreakers, Oden and Frej. During the test, helicopter images and an onboard camera were utilised to document the parallel channel fracturing events. With the collected data, we strive to quantify if there is a prominent relationship between parallel channel spacing and the corresponding managed ice floe size. In order to analyse the floe size distribution and its relationship with channel spacing from helicopter images, we developed an image segmentation method that propagates visually identifiable seeding cracks in the image. In addition, onboard camera images were utilised to yield the frequency of parallel channel fracturing events. Given the ice conditions and Oden's specific structural form, with all the different channel spacing tested, it turned out that a channel spacing over 200 m would already prohibit the development of parallel channel fracturing events. Most of the observed events take place when the spacing is smaller than around 100 m. In addition, as was expected, more frequent fractures are taking place with narrower channel spacing, e.g., distances smaller than 30 m. The relationship between managed ice floe size and channel spacing are studied. It is found that almost all (100%) of the produced downstream floe sizes are smaller than twice the channel spacing; 90% of them are smaller than 1.5 times of the spacing; and the majority of them (from 46% to 80%, depending on the spacing distance) are smaller than 1 time of the channel spacing. With such quantified relationships, we can practically estimate the size of the managed ice floes based on known/expected channel spacing.
A Shipborne Measurement System to Acquire Sea Ice Thickness and Concentration at Engineering Scale
Lu, Wenjun (Norwegian University of Science and Technology) | Zhang, Qin (Norwegian University of Science and Technology) | Lubbad, Raed (Norwegian University of Science and Technology) | Løset, Sveinung (Norwegian University of Science and Technology) | Skjetne, Roger (Norwegian University of Science and Technology)
Abstract Sea ice concentration and thickness are important parameters for the calculations of ice actions and their effects on Arctic offshore structures and for the evaluation of icebreaker performance. Various methods exist nowadays to monitor these parameters, ranging from geophysical scale to local scale. During the Oden Arctic Technology Research Cruise 2015 (OATRC’ 15), we installed both Ice Concentration and Ice Thickness cameras and developed corresponding algorithms to achieve real time quantification of ice concentration and visual estimation of ice thickness information. For the ice concentration analysis, we utilized both the global Otsu method to categorize an image into two regions (black water and white ice); and the K-means method to identify more regions based on the gray scale from the image. With the methods, we conducted a case study by analyzing the ice concentration in a selected time window. In the case study, we include both dry ice (in white color) and wet ice (in gray color, generally composed of ice rubbles, young ice, and melt ponds) as ice region for the K-means method. The K-means method yields higher ice concentration values in comparison to the global Otsu method, in which, melt ponds/young ice was frequently mistaken as open water. It turns out that the K-means method enables more flexibility to cope with the complicated ice environment by separating the image into more regions that can be included as ice in an ice concentration analysis. For the ice thickness camera, the intention was to capture the events while a broken ice piece is tilted, next to the ship side, and expose its thickness region to the camera. In this paper, we developed an automatic tracking algorithm to sift these events out from all the images taken by the Ice Thickness acquisition system. After projecting a grid with physical length onto the image, the ice thickness information can be visually quantified. We compared the ice thickness obtained from the Ice Thickness camera and that obtained by an Electro-Magnetic inductive device in a selected time window. The results agree well with each other. Considering the advantages and disadvantages of each method, this demonstrates the benefits of combining redundant approaches for obtaining the ice thickness information with a higher degree of confidence.