Synergistic applications of autonomous underwater vehicles and regional ocean modeling system in coastal ocean forecasting

The potential for using synergistic combinations of measurements from autonomous underwater vehicles (AUVs) and output from three-dimensional numerical models for studying the central California coastal region is demonstrated. Two case studies are used to illustrate the approach. In the first, propeller-driven AUV observations revealed a subsurface salinity minimum in northern Monterey Bay. A Regional Ocean Modeling System (ROMS) reanalysis of the three-dimensional flow in the region suggested an offshore source for this water and particular propagation pathways from the south and west into the bay. In the second case study, the effectiveness of assimilating observations in improving the ROMS reanalysis fields is investigated. A significant improvement, especially in the salinity fields, is demonstrated through a single glider deployed outside the intensive observational domain. These results suggest that investigation of more sophisticated techniques for using data and models together is warranted. Such techniques include increasing model resolution in areas of interest identified by observing platforms and using model-based ‘‘targeted observing’’ techniques to identify areas of uncertainty in the flow to guide placement of observational assets. Observations that adequately sample the three-dimenTraditional oceanographic measurements include either in sional (3D) structure of the coastal ocean are very sparse. situ or satellite observations. In situ measurements are mostly one dimensional (1D) at a single point (e.g., drifters or moorings) or two dimensional (2D) through a cross 1 Corresponding author ([email protected]). section (a ship track). Satellite instruments can provide a Acknowledgments greatly increased number of observations and coverage, but The research described in this paper was carried out, in part, at their measurements still provide a 2D view of the ocean the Jet Propulsion Laboratory (JPL), California Institute of surface. Using these 1D or 2D observations, we can only Technology, under contract with the National Aeronautics and infer or speculate on the complex 3D structure of the ocean. Space Administration. Computing support from the JPL Supercomputing Project is acknowledged. Thanks to James C. Recent development of Autonomous and Lagrangian McWilliams and his team for their help with the model. The Platforms and Sensors (ALPS) opens a new opportunity to contributions of Peggy Li, Kyungjeen Park, Quoc Vu, Xiaochun observe the 3D structure of the ocean. Over a limited area Wang, and Carrie Zhang to the model operation during and after of interest, a fleet of ALPS (e.g., gliders) has the potential the 2003 and 2006 Monterey Bay field experiments are acknowl­ of mapping this 3D structure on synoptic time scales. The edged. Thanks to Shelley Blackwell and Ian Robbins at California Autonomous Ocean Sampling Network (AOSN) field Polytechnic State University at San Luis Obispo for providing experiment (Curtin et al. 1993) conducted in the Monterey support for autonomous underwater vehicle deployments and Bay, California, during August 2000 demonstrated the glider deployments, respectively. The original high-frequency feasibility of glider technology in making routine oceano­ radar data were supplied by Jeffrey Paduan at the Naval Postgraduate School. The M1 mooring data were provided by graphic measurements. A follow-on field experiment, Francisco Chavez at the Monterey Bay Aquarium Research known as AOSN-II, deployed a fleet of more than a dozen Institute. gliders in the Monterey Bay in an attempt to monitor and