MATCHED FIELD PARAMETER SUBSPACE METHODS FOR SELECTIVE HIGH RESOLUTION ESTIMATION WITH REDUCED SENSITIVITY TO MISMATCH
HA CHANDLER, Christopher Feuillade, GB SMITH · 2024
Sector-focused estimation applied to matched-field localization has been shown to effectively enhance range-depth resolution compared with conventional Bartlett estimation.while remaining insensitive to environmental parameter mismatch.The insensitivity to mismatch is evident in parameter space ambiguity surfaces.where the response is typically broad with low resolution.For source localization purposes this is desirable.However, for environmental inversion, high resolution in parameter space is usually required, with a corresponding insensitivity to measurement errors.This may be accomplished using sector-focusing methods by including, in the search sector.replicas derived from a set of "neighboring" environmental parameters.This study demonstrates the use of these methods to increase resolution in environmental parameter space, while mitigating against measurement uncertainty in other parameters.