Towards Improved Ocean State Estimation

Abstract

Abstract : Physical oceanography has moved into a new era, one in which there is a relative wealth of data from the Global Ocean Observing System. This system is comprised of both in situ instruments, such as Argo profiling floats, drogued drifters and moored arrays, and remote sensing satellites, such as altimeters and radiometers. Each of the datasets is valuable, yet each suffers inherent deficiencies. To make the most of these data, they need to be combined through optimal interpolation and via assimilation into dynamical models. These statistical analyses are necessary steps in their quantitative use. A key ingredient of the statistical interpolation is the covariance between observable ocean state variables and the ocean circulation. The focus of this proposed research was to gain a better understanding of these multivariate covariances through the use of output from a global, eddy-resolving ocean general circulation model simulation. The result of the work will be better estimates of the covariance functions and from these, improved analyses of the ocean state and enhanced forecasting capabilities.