Geophysical support for aqueous fluids in the deep crust: seismic and electrical relationships

Abstract

Previous studies have shown that up to a few per cent porosity filled with saline fluid in the lower crust can explain many of the regions with: (1) low electrical resistivities, (2) velocities that appear to be too low for the otherwise inferred mafic composition, and (3) strong lower crustal reflectivity. Several predictions of the free porosity model are examined in this article. A compilation of approximately coincident magnetotelluric electrical resistivity and refraction seismic velocity data for the lower continental crust is presented to test the predicted correlation. In spite of the limited geographically coincident data and the difficulties of ensuring accurate depth coincidence and of anisotropy effects, there is a general trend of decreasing velocity with decreasing resistivity. The data are scattered, but most fall between the reasonable bounds provided by pore geometry models with effective aspect ratio (for velocity) and Archie's Law pore tortuosity exponent (for resistivity) pairs of 0.03 : 2.0 and 0.1: 1.5 respectively. As in previous compilations, shield areas tend to have both higher resistivities and higher velocities in the lower crust compared to Phanerozoic areas, although there is overlap for both parameters. A general correlation is also found between the top of low resistivity layers and the top of lower crustal reflective zones with the 400-450°C isotherms. Possible explanations of this correlation with temperature include (1) an association with the brittle-ductile transition, below which pore geometries are such as to hold fluid in the required configuration, and (2) control provided by metamorphic reactions that restrict free fluid to below this depth. To constrain better the pore geometry, a compilation of the limited data on lower crustal Poisson's ratio shows most values ∼0.28. This is consistent with a mainly mafic composition with up to several per cent porosity. Reasonable pore geometry distributions predict a small decrease or constant Poisson's ratio with increasing porosity. While each of the three lower crustal geophysical data types have other reasonable explanations, the apparent correlations above provide support for the fluid-filled pores in the lower crust. The problems of the low permeability required to keep fluid in the lower crust, and of pore fluid consumption in retrograde metamorphic reactions during cooling are discussed briefly. Two mechanisms are suggested as means of producing a low-permeability cap in the middle to deep crust: one invokes deformation of textural equilibrium pore geometries by small deviatoric stresses, the other lower crustal shear processes. There remains some difficulty in reconciling free aqueous fluids in the lower crust with the expected retrograde metamorphism that should take up water into hydrated mineral assemblages.