Deep ocean sound speed profile estimation using matched autoproduct processing

Abstract

Sound speed profile (SSP) estimation from acoustic measurements alone in ocean environments is often a highly undetermined problem. Many methods have been developed which match arrival times or acoustic pressure observations to estimate a reduced-order SSP described with shape functions or empirical orthogonal functions (EOFs). These EOFs are commonly derived from multiple in situ measurements. This presentation describes an alternative method that matches autoproduct fields to invert for deep ocean SSPs without knowledge of the source waveform. The inverted SSPs are parameterized with EOFs derived from synthetic SSPs extracted from HYCOM (hybrid coordinate ocean model). This method is developed and illustrated using simulated long-range deep-ocean propagation data for frequencies in the 100s of Hz. The technique is also assessed with measured propagation data from the PhilSea10 experiment. [Sponsored by ONR]