Design of Field Experiments for Adaptive Sampling of the Ocean with Autonomous Vehicles

Due to the highly non-linear and dynamical nature of oceanic phenomena, the predictive capability of various ocean models depends on the availability of operational data. A practical method to improve the accuracy of the ocean forecast is to use a data assimilation methodology to combine in-situ measured and remotely acquired data with numerical forecast models of the physical environment. Autonomous surface and underwater vehicles with various sensors are economic and efficient tools for exploring and sampling the ocean for data assimilation; however there is an energy limitation to such vehicles, and thus effective resource allocation for adaptive sampling is required to optimize the efficiency of exploration. In this paper, we use physical oceanography forecasts of the coastal zone of Singapore for the design of a set of field experiments to acquire useful data for model calibration and data assimilation. The design process of our experiments relied on the oceanography forecast including the current speed, its gradient, and vorticity in a given region of interest for which permits for field experiments could be obtained and for time intervals that correspond to strong tidal currents. Based on these maps, resources available to our experimental team, including Autonomous Surface Craft (ASC) are allocated so as to capture the oceanic features that result from jets and vortices behind bluff bodies (e.g., islands) in the tidal current. Results are summarized from this resource allocation process and field experiments conducted in January 2009.
Singapore. National Research Foundation