A Random Field Framework for Boundary Conditions in Numerical Models of Coastal and Ocean Circulation

University Of Miami, Coral Gables FL

Investigators

Abstract

0136700/Chin This project introduces a novel approach to representing boundary conditions in ocean models. The idea is, absent the ability to follow small-scale physical processes in detail, to represent the influence of these small-scale processes on the larger-scale dynamics resolved by the model by a stochastic process. The project will try several different auto-regressive representations of the boundary conditions and examine how the large-scale structure of the flow reacts to these. The model results will be compared with data from observations near the Florida Current made with HF radar. Using inverse techniques, based on a Markov random field approach, the values of parameters in the stochastic models of boundary conditions will be estimated. The sensitivity of these inversions will also be determined. In larger domains, dynamically similar to the North Atlantic, the question of whether stochastic boundary conditions along the eastern boundary can lead to internal jets (beta plumes) will be studied. Initially, the numerical circulation model will be a reduced-gravity, quasigeostrophic model in idealized geometries. The most effective approaches to stochastic boundary conditions will then be transferred to a primitive-equation OGCM.