Michael J. Longfritz
William L. Siegmann
Melvin J. Jacobson
Rensselaer Polytechnic Inst., Troy, NY 12180-3590
Mohsen Badiey
Ocean Acoust. Lab., University of Delaware, Newark, DE 19716
The variances of horizontal acoustic wave numbers in stochastic
shallow-water environments are estimated using empirical orthogonal functions
(EOFs). Using perturbation methods and an adiabatic normal-mode propagation
model, the wave number fluctuations are related to range variations in the
ocean and/or sediment sound-speed profiles, which can be conveniently and
efficiently represented by EOFs. This relationship can be used to estimate both
the wave number deviations arising from selected realizations of the stochastic
ensemble and the overall wave number variance. The procedure is illustrated by
application to a shallow-water waveguide with a multilayered sediment bottom.
Both the layer depths and the intralayer sound speeds are modeled as random
variables. The accuracy of the estimation procedure is investigated for various
choices of the layer depth and sound-speed statistics by comparing with results
from computational simulations. A particular environment examined is the New
Jersey Shelf Atlantic Generating Site, where relatively extensive geoacoustic
profiles are available [M. Badiey et al.,