The Interference, Diffraction and Scattering of Acoustic Waves in a Turbulent Atmosphere.

The existing theory to predict the interference of sound waves above a hard boundary allows for partial coherence between direct and ground-reflected sound but is limited by the need for one adjustable parameter. The theory is reexamined to include, in addition to the transverse autocorrelation, the longitudinal autocorrelation of the interfering waves but, when recompared with previous measurements over asphalt, is still limited by the need for the same adjustable parameter. The theory is further extended to include propagation over ground of finite acoustic impedance for both short and longer distances of propagation where, in addition to phase covariance, covariance of sound amplitude is important while phase-amplitude covariance can always be neglected. The new theory provides an explanation for measurements of jet noise over grass where the sound pressure levels are up to 10 dB higher than predicted by coherent acoustic theory. Measurements of traffic noise attenuated by a barrier alongside a highway are described. They show two characteristic discrepancies when compared with classical diffraction theory. To investigate these discrepancies an idealised experiment was performed using a point source and a model barrier erected on a flat asphalt surface. The sound pressure levels measured behind the barrier are, when compared with two formulations of classical diffraction theory, up to 10 dB higher than predicted. An empirical correction to the classical line-integral theory, as well as an alternative diffraction theory, show better agreement with the measured results but does not resolve the other discrepancy in sound levels that occurs mainly at the higher frequencies. A calculation is described to estimate the amount of sound energy scattered by atmospheric turbulence above the barrier edge. The sound pressure levels calculated from the logarithmic sum of the scattered energy and the energy from diffraction at the barrier edge show better agreement with the results than just diffraction theory alone, especially at the larger distances.