Numerical methods to model infrasonic propagation through realistic specifications of the atmosphere

Abstract

To model infrasonic propagation, two areas must be adequately addressed. First, the environment must encompass a domain spanning from the ground to the lower reaches of the thermosphere, and its properties must be resolved at a scale comparable to the acoustic wavelengths of interest. Second, the relevant fundamental physics that influence the wave propagation must be captured in the numerical models that are applied. To address these challenges, a large amount of work has been completed by researchers in fields spanning from traditional acoustics to upper atmospheric physics. In this chapter, this body of work is reviewed with an emphasis on its applicability to the modeling and interpretation of infrasonic observations. Global climatological models have largely been replaced in current infrasound modeling practice by atmospheric specifications that combine output from numerical weather prediction models for the lower atmosphere with empirical models for the upper atmosphere. Recently developed specifications incorporate higher-resolution, regional or mesoscale weather analysis products in order to improve fidelity below altitudes of 50 km. Specifications that utilize a terrain-following coordinate system, including high-resolution topography, enable the incorporation of additional physical effects. These specifications are unable to resolve all fine-scale stochastic phenomena, e.g., atmospheric irregularities smaller than the model resolution, fine-scale structures at altitudes above 50 km, and gravity wave fluctuations. In particular, gravity waves are of interest because their spatial scales are of the same order as infrasonic wavelengths. Because atmospheric fine-scale structure is inherently turbulent, spectral formulations are used to capture the energy distribution across space-time scales relevant to infrasound and to generate representative realizations of variable fields. The fundamental physical processes that affect infrasound, much like those of higher frequency acoustics, include refraction, diffraction, scattering, absorption, dispersion, and terrain. In addition, nonlinear effects can become significant near the source and at high altitudes. Several numerical approaches have been used for modeling infrasonic propagation often based on analogous models for underwater and outdoor sound propagation. They can be loosely classified under the headings of geometric (ray tracing, Gaussian beam, tau-p), continuous wave (parabolic equation or PE), full-wave (normal mode, finite-difference time domain, time-domain PE), and nonlinear (e.g., NPE). A good introductory source to the underlying computational aspects is (Jensen F, Kuperman W, Porter M, Schmidt H (1994) Computational ocean acoustics. AIP Press, New York). (Although this reference is primarily intended for ocean acoustics, many of the models carry over to the atmosphere and the presentation is superb.) A general overview of the propagation models is presented here with their corresponding strengths and weaknesses highlighted in the context of infrasonic studies. Also included are examples of modeling studies using high-resolution atmospheric specifications. © 2009 Springer Science+Business Media B.V.