Ultrasonic velocity measurements in volcanic rocks: correlation with microtexture

SUMMARY This paper summarizes results of a study of porosity, permeability, microstructure and acoustic properties of volcanic and pyroclastic rocks from Campi Flegrei (CF) and Mt. Etna (ET), Italy. We have measured the hydraulic, transport and acoustics properties of 28 room-dry samples at ambient conditions, 25 room-dry samples under confining pressure (up to 60 MPa) and 5 brine saturated samples under pressure (up to 45 MPa effective pressure). We established the following range of porosity, permeability, and Vp and Vs variations as functions of mineralogy and differential pressure in CF and ET lithologies: 1 Porosity CF pyroclastic rocks 30‐60 per cent CF and ET lava rocks 2‐20 per cent 2 Permeability CF pyroclastic rocks 10‐1000 mD CF and ET lava rocks 0.01‐100 mD 3 Velocity and Quality Factor CF pyroclastic rocks Vp 2‐3 km s x1 Vs 1‐2 km s x1 Qp 5‐80 CF and ET lava rocks Vp 3.5‐5.5 km s x1 Vs 2‐3 km s x1 Qp 10‐115 Mineralogy and microstructure govern the acoustic and petrophysical properties of these rocks under pressure. In pyroclastic rocks, changes in acoustic response are directly related to presence of zeolites and pumice and their reactions with the pore fluid. In dry conditions, collapse of the internal texture of pumice leads to decreasing velocity with pressure. In saturated conditions, the water‐zeolite interactions compete with effects due to collapsing internal textures and so velocity change with pressure is not as pronounced. The microstructural changes were confirmed by analyses of optical, hydraulic and transport properties after pressure: CT-scans show a macroscopically more compact structure; under optical microscopy, pumice and zeolitized pumice appear shredded. A significant reduction in porosity and permeability is measured after pressurization. In lava rock samples, acoustic velocities increase in function of pressure and the velocity‐pressure relationships are characteristic of samples with rounded pores. Our results emphasize the importance of conducting velocity measurements at simulated in situ conditions. By constraining the computations using site- and depth- specific rock physics properties, differences between ground deformation models in volcanic areas can be assessed and predicted more reliably thus reducing volcanic hazard