Abstract

The theory of linear poroelasticity developed by Biot [4] in the lowfrequency limit can be profitably used to estimate porosity in a fluid-saturated continuum from measured transversal (shear) and longitudinal (compression) wave velocities. Porosity is an important state parameter which controls the hydromechanical response of porous materials such as soils. While in fine-grained geomaterials such as silts and clays porosity can be easily measured in laboratory on undisturbed samples, in coarse-grained soils such as sands and gravels undisturbed sampling is difficult and the possibility of estimating porosity from direct inversion of seismic velocities, which are routinely measured using geophysical seismic prospecting methods, is appealing. The methodology has been proposed and successfully applied by Foti et al. , 2002 [13] and it is based on the apparent insensitivity of the inversion procedure on the Poisson ratio of the (evacuated) soil skeleton which is a-priori an unknown quantity. This paper attempts to thoroughly investigate the stability of the inversion of measured seismic wave velocities for porosity estimation and to assess the degree of well-posedness of the procedure.