Mirages in shallow water matched field processing

Abstract

Broadband matched field processing (MFP)-derived estimates of 3D source location using data from hydrophone line arrays deployed in various geometries, i.e., vertical, horizontal, and tilted 45° from vertical, are presented in this paper. These data were collected in two shallow water (100–200 m) experiments off the coast of San Diego. Results show that estimates of source range and depth remain surprisingly coherent in the presence of large mismatch in bathymetry, but are offset from the true position by as much as 100%. The offsets are independent of array geometry. In contrast, for estimates of source azimuth, bathymetry mismatch typically causes a degradation in MFP correlation rather than an appreciable offset. However, errors in the assumed tilt from vertical of an array can lead to large offsets in the estimated source azimuth, particularly as the nominal angle of the array from vertical becomes smaller. Predictions from a simple analytical model based on adiabatic normal modes in ideal waveguides provide good fits to the broadband MFP results.

Interferometric array imaging in clutter

“…This is the form of a matched field functional [1,18,28,35], especially when the exponential of the travel time is replaced by the deterministic Green function…”

On the Feasibility of a Matched-Field Inversion in a Three-Dimensional Oceanic Environment Ignoring Out-of-Plane Propagation

“…In addition, the TL comparisons are still excellent and the Bartlett values at the MAP estimate is unchanged. In other words, the result at the range of 5 km is a typical mismatch case [26] where, in addition, we have no indication that the estimates are erroneous. In the context of real data applications, in which the true values Fig.…”

Statistically stable ultrasonic imaging in random media

“…Two well-known frequency-domain imaging methods, matched-field processing (Bucker, 1976;Baggeroer et al, 1988;Baggeroer et al, 1993;Jensen et al, 1994;D'Spain et al, 1999) and MUSIC (Schmidt, 1979;Johnson, 1982;Johnson and DeGraaf, 1982;Schmidt, 1986;Stoica and Nehoral, 1989;Stoica and Moses, 1997;Devaney, 2002), will be presented next to show how these methods behave in the types of random media we are considering.…”