Polynomial chaos expansions for travel time and amplitude uncertainty in acoustic time series in ocean waveguides

Polynomial chaos (PC) techniques have shown themselves to be a powerful technique for the estimation of the statistical properties of ocean acoustic propagation for situations where the vertical and horizontal structure of sound speed uncertainty can be specified in terms of empirical orthogonal functions (EOFs) with known amplitude probability distribution functions. Here we extend the CW solutions for the statistical properties of the normal models describing the acoustic field in the presence of internal waves with Gaussian distributed EOF descriptions to the time domain to describe the scintillation of the acoustic time series. Results are compared to exact solutions as well as to estimates based on the adiabatic approximation. Results show that the PC solutions do a much better job of capturing the true variability of the exact solutions for the time series variability than the adiabatic approximations does. [Work supported by ONR.]