Remote sensing of sea state for ships navigating in arctic waters

The United States has a growing interest in navigating the arctic waters Greenert [6]. The realities of global warming are manifesting in more navigable waters in the arctic, creating growing concerns of national security and increased interest in the economic potential of the arctic. As such, ensuring safe sea travel for vessels in harsh arctic conditions is now a primary concern. In order to achieve reliable sea-keeping to minimize the risk of capsizing, I am interested in developing a framework to address the potential scenario of ice accumulation on a ship vessel, thus changing the internal mass properties the vessel was designed for or would have without ice accumulation. Previous studies have indicated that the main agent for icing is from sea-spray due to ship motion interacting with ocean waves [4, 11, 12]. It is, therefore, critical to know both the internal mass properties of the ship, as well as the incoming wave or sea state that is providing an excitation to the ship. The effects of the incoming waves are two-fold: the excitations bear influence on the resulting sea-spray, and the excitations also pose a potential risk for the ship to lose sea-keeping and stability capabilities. The former is a difficult problem to study since the data are difficult to collect in a controlled environment. Many models have been created in order to approximate this, from statistical models to physics-based models. If we take the information from these models as to where like is likely to form, we can focus on the latter and study only the incoming waves and the resulting ship motion, using the ship’s inertial measurement unit. Given that we can properly describe the sea state (wave height, period, direction and velocity of wave), we could compare the actual motion of the ship to a simulated fluid-structure interaction model of the ship in a virtual sea. In this study, I am interested in investigating the feasibility of recovering the following to use in the described framework: (1) the current sea state in the area surrounding a vessel potentially travellign in the arctic, including wave height and period and (2) the direction and velocity of the waves. There are several related applications that may provide insight for this project. Concerns about data availability and sea state visibility will be critical. The remote sensing instrument would ideally provide data in high temporal frequency and be made available in near-real-time from the time of collection, and must be readily collected even with significant cloud cover or other obscurities that may occur in the arctic. These preliminary sources indicate that the problem I am approaching is likely to be technologically feasible. The following is a brief synopsis of different remote sensing techniques that have been used to infer sea states. There are methods to determine the scale and sea state from a standard video, independent of calibration [13, 14]. This method relies on computer vision techniques, along with prior knowledge of wave dynamics and physics, to infer the wave heights and velocities without any formal calibration stage, making this unique from most other techniques. Resolution is in the correct scale, with the studies ranging from 0.1-0.7 meters per pixel. While this study has shown promising results, the use of a normal video is extremely limiting. It is required that the system work both in night time as well as in low visibility scenarios, such as heavy fog. The logistics of directionality would also have to be considered in order to obtain a panoramic view. Existing data on ocean wind vectors, a parameter closely tied to sea state (a modern version of the Beaufort scale, which relates wind speeds to form and height of waves, along with a qualitative