A comparison of sediment reflection coefficient measurements to elastic and poro-elastic models

Abstract

This work compared three plane wave reflection coefficient models to laboratory sediment reflection measurements collected using a spherical source and receiver. Plane wave decomposition was used to modify the reflection coefficient models to include the inherent spherical effects such as lateral wave interference for direct comparison with the data. Sets of data at two transducer separation distances (1.27 and 0.25m) were collected to compare the range dependence of the spherical effects. Bandpass filtered linear frequency-modulated chirps from 30 to 160kHz were used to measure frequency dependence. Grazing angles from 5° to 75° were measured to compare angle dependence. Each set of data was collected along approximately 3m of smoothed sediment for spatial averaging. Unavoidable experimental effects including transducer response, beam pattern, and spherical spreading were accounted for in order to compare the reflection coefficient measurements with the modified models. Significant spherical wave effects were measured in the data. Three reflection coefficient models were considered: the viscoelastic model, the grain shearing model [M. Buckingham, J. Acoust. Soc. Am. 108, 1796–2815 (2000)] and the effective density fluid model [K. Williams, J. Acoust. Soc. Am. 110, 2276–2281 (2001)]. The viscoelastic and grain shearing models predicted values for the reflection coefficient that were not within the 95% confidence interval for low frequencies. The data exhibited high variance which was frequency and angle dependent. This variance is not likely to be caused by variations in bulk properties as defined by the fluid or viscoelastic models. The cause of the variance will be considered in subsequent publications.