Estimation of porosity and water saturation in dual-porosity pyroclastic deposits from joint analysis of compression, shear, and electromagnetic velocities

Abstract

In situ measurements of porosity and water saturation of pyroclastic deposits have the potential to improve interpretations of geology and hydrology in volcanic regions, and to provide more accurate estimates of dense rock equivalent for volcanic eruptions. However, rock-property models must consider the dual-porosity structure of pyroclastic deposits (i.e., vesicles within pumices and intergranular pores). Vesicularity, intergranular porosity, and water saturation all affect the density, elasticity, and dielectric properties of pyroclastic materials, which control seismic and electromagnetic velocities. The data from active seismic and ground-penetrating radar (GPR) techniques may improve porosity and water saturation estimation if the responses of seismic and electromagnetic velocities to porosity and water saturation variations are complementary in pyroclastic deposits. We developed a dual-porosity petrophysical model to calculate seismic and electromagnetic velocities of pyroclastic deposits with known intergranular porosity, vesicularity, and water saturation, and we tested our ability to estimate porosity and water saturation from field measurements of seismic and electromagnetic velocities in pyroclastic deposits at Mount St. Helens, Washington, USA. Our petrophysical model demonstrates that seismic velocities are more sensitive to intergranular porosity and less sensitive to vesicularity; electromagnetic velocity is primarily controlled by volumetric water content. In a multioffset GPR and active seismic case study, seismic first-arrival traveltime tomography and multichannel analysis of surface waves can resolve high-velocity anomalies caused by porosity reduction. Joint petrophysical inversion of electromagnetic and seismic velocities indicates that although intergranular porosity and water saturation are well-constrained (i.e., standard deviations of approximately 0.05), quantitative estimates of vesicularity remain less certain (i.e., standard deviation of approximately 0.21), due to weak sensitivity.