Model Assessment and Deployment Strategies for Drifting Instruments

Abstract

LONG-TERM GOALSThis is part of a collaborative program Optimal Deployment of Drifting Acoustic Sensors, ODDAS, which involves academic, private sector, and Navy scientists. The long-term goal of ODDAS is to develop strategies for the deployment of drifting sensors that maximize the amount of environmental information collected with the fewest sensors. The primary long-term goal of our component is to quantify oceanic submesoscale stirring and advective transport processes and to assess their roles in the deployment strategies. The secondary goal is to develop novel and quantitative assessments of the ability of predictive models and real-time observations to determine the movement of sensors and the location of material boundaries or advective pathways in the ocean. OBJECTIVESTwo objectives were pursued during FY06:• Locate hyperbolic trajectories and calculate associated inflowing and outflowing manifolds in EAS16 hindcasts for the region around Taiwan.• Assess the ability of EAS16 to predict the movement of drifting sensors. APPROACHAs documented in publications and previous progress reports we have developed a variety of Lagrangian methods that quantify advective processes from synoptic velocity archives. These methods are used in this project to configure deployment schemes of drifting sensors so as to maximize data return from drifting sensor arrays with a minimum number of units. This requires meeting three diverse, and occassionally contradictory, user requirements. One is to maximize the spatial coverage for regions of interest. Another requirement is maintain array coherence for a specified time period. The third requirement is to sample denied areas. To meet these requirements it is also important to decide whether to deploy all sensors at one time or to use a time staggered strategy.The key of our approach is the capability to locate critical regions from the synoptic velocity archives. These are usually found in regions where the velocity is small. However, small velocity regions in the