When an airborne acoustic wave is incident at the ground surface, energy is coupled into the ground as seismic motion. In a previous publication [Sabatier et al., J. Acoust. Soc. Am. 78, 1345–1352 (1986)] the ground surface was modeled as an air-filled poroelastic layer overlying a semi-infinite, nonporous elastic substrate. In this work, the model is extended to include calculations of the normal seismic transfer function (ratio of the normal soil particle velocity at a depth d to the acoustic pressure at the surface). Measurements of the seismic transfer function for three sites are considered and compared to the predicted values. Generally good agreement between theory and experiment is achieved by best fits assuming the soil or seismic attenuation. This is accomplished by specifying the ratio of the imaginary to real part of the measured seismic p- and s-wave speeds. The seismic transfer functions quite typically exhibit minima and maxima which are associated with the seismic layering of the ground surface. Typical layer depths are 1–2 m. An analytical expression predicting the location of these maxima is offered based on hard substrate and the experimental and theoretical comparisons are reasonable.

The surface of the ground is modeled as that of an air-filled poroelastic soil layer of known thickness overlying a semi-infinite nonporous elastic substrate. Using a modified form of the Biot–Stoll differential equations for wave propagation in fluid-saturated porous media, propagation constants for the two possible dilatational waves and the shear wave in the poroelastic layer are determined. The dilatational waves are identified as a fast wave, moving predominately in the solid frame, and a slow wave, moving predominately in the pore air. The elastic moduli in the substrate are assumed to be those of the solid grains of which the poroelastic soil layer is composed. Intergranular friction in the soil and substrate is assumed to be negligible. Boundary conditions at the air–soil interface and at the porous soil–substrate interface are applied to determine, numerically, the displacement amplitudes of the allowed wave motions. From the incident and reflected amplitudes at the air–soil interface, the normalized ground surface impedance is calculated as a function of angle of incidence and of frequency. In this paper, the response of the pore fluid and frame to airborne acoustic waves is considered and those ideas will be pursued in a later publication. The predicted impedance at normal incidence is compared with measurements of the impedance of a sandy soil for which measurements of the various parameters required by the theory are also available. The predictions of impedance are found to be in tolerable agreement both with measured data and with predictions of a simpler model of the surface as that of a rigid porous semi-infinite homogeneous medium. Calculations of the surface impedance as a function of angle of incidence suggest that the porous medium is locally reacting.

Simultaneous measurements have been made of sound pressure above the ground and seismic velocity below the ground surface resulting from a source suspended in the air a variable distance from the surface. The ratio of seismic velocity to acoustic sound pressure has been determined; there are peaks in the ratio in the vicinity of 45 and 90 Hz depending on the height of the speaker. The source-receiver distance was 10, 30, and 60 m; the source height was varied between 1 and 10 m. The frequency of maximum acoustic coupling was found to agree well with theory; the first and third shear modes appear to be excited. Results for vertical, horizontal, and radial motion indicate the coupled seismic signal is greatest for vertical, next greatest for radial, and least for transverse, though the difference between radial and vertical displacement velocities was not great and depended on the speaker altitude.

The rise times and overpressures for shock waves from supersonic projectiles have been measured under a variety of atmospheric conditions. These measurements extend the overpressure for which weak shock theory adequately predicts measured values to 3 kPa. Predicted shock rise times agree reasonably well with theory for shock overpressures between 0.04 and 0.5 kPa with differences increasing at higher overpressures. At overpressures greater than 0.5 kPa, measured rise times might be limited by the experimental apparatus. No correlation between measured rise times and atmospheric turbulence was observed.

A model for the response of a poroelastic layered soil to an incident plane wave developed in a previous paper [Sabatier et al., J. Acoust. Soc. Am. 79, 1345–1352 (1986)] is used to predict the complex sound pressure within the upper poroelastic layer. The predictions both of phase velocity and attenuation of the slow wave associated primarily with propagation in the pore fluid are compared with measurements made with a specially constructed probe microphone. The agreement between theory and experiment is good. The predictions for a layered poroelastic soil model are compared numerically with those of a semi-infinite rigid porous soil model and are found to differ only at frequencies higher than 1000 Hz. Analysis of the sensitivity to the theoretically predicted propagation constants in the poroelastic soil to the assumed value for the bulk rigidity modulus of the soil predicts that over the known range of rigidity moduli for soils it is possible to obtain a switchover between fast and slow propagation modes. This switchover occurs at the lower end of the possible range of values of the shear modulus. It is suggested that probe microphone measurements in air-filled soils offer a way of measuring flow resistivity and of deducing the structural parameters required for application of the Biot–Stoll model to water-saturated sediments.