Variability due to short-distance favorable sound propagation and its consequences for immission assessment

The specific noise immission from an (industrial) noise source is commonly assessed by short-term measurements. Good practice prescribes measuring under downwind conditions at modest wind speeds. Nevertheless, this still leads to large variation, even at short distances and needs quantification. More specifically, the variation in sound propagation due to the changing refractive state of the atmosphere and the relatively large variation in soil impedance one can find for (visually determined) "grassland" is studied. Highly detailed meteorological tower data were combined with measured grassland impedances. These data are fed to the full-wave one-directional Green's function parabolic equation sound propagation model. The variation, even under these good-practice measurement conditions, is found to be large, and strongly dependent on sound frequency, source height, receiver height, and propagation distance. When assessing the specific sound pressure level from a multitude of sources, this variation strongly decreases compared to a low-height single source. Besides absolute variations, fluctuations in the transmission loss between a close point and a more distant one are discussed in this paper. The variation ranges give an idea on this systematic uncertainty when performing short-term measurements, and their impact on convergence to yearly averaged equivalent sound pressure levels.