MODELING CORAL REEF ACOUSTIC PROPAGATION LOSS

Abstract

Quantification of an acoustic metric relating sound propagation on a coral reef to ecological parameters could enhance long-term ecological studies on coral reefs. The acoustic metric constructed in this study is coral reef propagation loss, or the sound intensity reduction with range, in a coral reef environment. Establishment of this benchmark parameterizes the geometric spreading and attenuation factors of the coral reef environment. Two analyses are conducted in which both passive and active acoustics are used to measure the sound propagation environment across different fringing coral reef sites. In the first analysis, two passive acoustic studies in the literature captured the ambient reef soundscape (abiotic and biotic factors within 0.1 - 5 kHz band) sound pressure level at various ranges offshore on six different coral reef sites worldwide. This broadband coral reef soundscape information is utilized to evaluate propagation loss. The second analysis is an active acoustic experiment conducted at two different Hawaiian coral reefs, where low-level tone transmissions (0.5, 2, 5, 10, 15 kHz) were recorded at various ranges offshore. These tones are evaluated regarding both frequency and range dependence of transmission loss. The development of the reef propagation loss metric is explored through application of existing literature models (cylindrical spreading, spherical spreading, Roger's Onboard Empirical Formula, Marsh-Schulkin, Extended Marsh-Schulkin, and Bellhop), as well as through nonlinear least squares inversion scheme models (sloped cylindrical spreading, spreading only, attenuation only, spreading and attenuation/Base Model) compared against the field data. Iterative minimization, performed with the nonlinear least squares method, provides indices of a geometric spreading factor and an attenuation factor of the specific environment. Root-mean-square error metric compared the accuracy of the multitude of models to the acoustic field data for both analyses. Results of the ambient coral reef soundscape study indicate geometric spreading factors a magnitude less than as predicted with conventional cylindrical spreading. Attenuation factors extracted correspond to dry silt sediment bottom values in the literature. Results of the transmitted tones study suggest nonlinear frequency dependence of the coral reef propagation environment. The field data indicated stronger contribution from spreading and minor contribution of attenuation in characterizing the transmission loss. Geometric spreading parameter estimates bracket that of cylindrical spreading. Broadband coral reef soundscape propagation exhibits significantly less propagation loss than a single tonal trans- mission within the same band and in a similar environment. These results further elucidate the complexity of the coastal environment and the frequency dependent nature