Bottom effects on matched‐field localization performance in the North Pacific

Abstract

PACIFIC ECHO was an experiment conducted jointly by the Naval Research Laboratory (NRL) and Defence Research Establishment, Pacific (DREP) in May–June 1986. Vertical array measurements of a 15‐Hz cw signal projected by the NRL MK‐VI source, towed at a depth of 100 m, were taken on the 675‐m‐long DREP array at both thinly (< 50 m) and thickly (> 500 m) sedimented areas. The array was located in separate cases in the direct path and the shadow zone, and the top sensor was at a depth of approximately 350 m. Matched field processing between theoretical and measured fields was applied to this data set in order to perform range‐depth localization. Both a linear (Bartlett) and a nonlinear, minimum‐variance (Capon) estimator were employed using theoretical fields generated by a normal modes program. The bottom model plays an important role in the shadow zone where acoustic energy is received predominantly via bottom interacting paths. The thin sediment site, in turn, is especially interesting because of the interaction with the elastic subbottom. The effect of differences in measurement site (thin or thick sediment), source‐received separation, and processor estimator on performance of the matched field processor are discussed.