Effects of strong sound velocity gradients on propagation of low-frequency impulse sound: Comparison of fast field program predictions and experimental data

As part of a large outdoor sound propagation experiment series, propagation of low-frequency impluse sound at ranges from 100 to 1400 m was studied at Haslemoen, Norway in February 1995. Sound sources were charges of 1 and 8 kg C4 explosives. Experiments were carried out in both a uniform forest and above a flat and uniform open field. Extensive meteorological measurements were carried out. Both automatic weather stations mounted in permanent towers and a tethered balloon were employed. The measurements resulted in a manifold set of sound velocity profiles. Occasionally, profiles were pure downwind or upwind, but mostly they were complex, often with large vertical gradients. One conclusion is that one should not uncritically use simple profile parameterizations, e.g., logarithmic, in sound propagation models. A fast field program (FFP), CAPROS, handles the complex sound-speed profiles well, predicting propagation of sound at 8, 16, 32.5, and 63 Hz. In this model, ground is considered a viscoelastic medium. At 63 Hz, the forest seems to cause an excess attenuation. At the lower frequencies, distinct effects of forest are not observed. There is no evidence that FFP predictions are less accurate in periods with high turbulence levels, which is here characterized with dynamical instability. However, variability of meteorological profiles causes predictions to be less accurate. This investigation suggests that it is sufficient to sample atmospheric variables up to 300 m above ground to predict sound propagation at the ranges studied here. (C) 1997 Acoustical Society of America.