Abstract:
This paper describes matched-field processing (MFP) of northeast Pacific Ocean data collected from a vertical line array (VLA) during PACIFIC SHELF 93. In this trial, sound at 45, 70, and 72 Hz from a source at a depth of 30 m towed downslope in water from 150 to 375 m deep was received on a 16-element VLA with sensors evenly spaced between 90 and 315 m. Using sparse bathymetric data, a single measured water sound-speed profile, and estimates of bottom properties, MFP correlations on individual ambiguity surfaces were found to be greater than 0.9 for the strongest signals. By applying an efficient linear tracker to the top 15 peaks extracted from each MFP ambiguity surface at a fixed depth, the source's depth and range as a function of time were recovered. In contrast, the source position could not be determined unambiguously from most of the ambiguity surfaces. These results show the reduction in ambiguity of source position through the use of an efficient tracker. Results are also presented that show the localization or detection performance improvement from reducing mismatch from array tilt, improving the tracking model, searching for the best bathymetry in an N by 2-D model, and taking mode coupling into account.