Ocean Environmental Inversion via Acoustic Data Assimilation

摘要

Modem methods for passive source localization, such as matched-field processing (MFP), require accurate knowledge of the ocean acoustic environment. This paper presents an approach of environmental inversion via acoustic data assimilation, which is intended to reconstruct the true environment parameters through iteratively reducing the environmental uncertainty relevant to sound propagation. The approach uses the same framework laid down in (Elisseeff 2002). In the context of sound speed profile (SSP) inversion, it exploits a variety of information sources including direct local sound speed measurements, an oceanographic model of the sound speed field, a full field acoustic propagation model, and full field measurements in a configuration similar to MFP. A few important improvements have been made in numerical implementations: 1) a more realistic shallow water model from the Shelf Break PRIMER 1996 experiment is considered, including the measured SSP statistics; 2) a set of end system-decoupled representation (system orthogonal function-SOF) is used to describe the perturbation of the SSP; 3) a nonlinear optimization, called Adaptive Simulated Annealing (ASA), is utilized to search the parameter landscape for the global minimum of the cost function. The results have further developed the approach in (Elisseeff 2002) and also for the first time the environmental inversion is coupled with MFP through environmental parameterization.