Coastal Ocean Observing System Modeling: Data Assimilation and Adaptive Sampling Design.

摘要

Accurate analyses of ocean currents, waves, temperature and salinity are required in order to forecast sediment and bio-optical properties in coastal waters that are critical to the performance of naval environmental sensors and operations. Analyses of these features can be significantly improved through the combination of observations and models by advanced data assimilation. With ONR support, the Regional Ocean Modeling System (ROMS) code has undergone extensive development and is being adopted by an expanding user community for ocean reanalysis and forecast applications. An important new enhancement in the modeling system is the formulation of tangent linear and adjoint codes and, built upon these, the capability for 4-dimensional variational data assimilation, sensitivity and stability analysis, and optimum generation of forecast ensembles to explore predictability limits. We will develop a coastal ocean modeling system that uses advanced data assimilation techniques with observations from new, rapidly deployable and relocatable coastal ocean observational assets to improve ocean forecasts in coastal regions that are of increasing interest to Navy operational requirements. Our system for regional ocean prediction will have the capability to use all available data from comprehensive coastal observational networks comprised of multiple CODAR installations, cabled observatories, autonomous gliders, and satellite imagery.