Geophysical Assessment of Migration and Storage Conditions of Fluids in the Earth’s Upper Mantle

Tuesday, 16 December 2014: 1:55 PM
Anne Pommier, University of California San Diego, La Jolla, CA, United States
An investigation of the relation between electrical and seismic parameters is presented for different tectonic contexts. By combining electrical laboratory data and field results (electromagnetic and seismic), constraints on melt storage conditions at depth can be improved. Electrical and elastic rock properties are sensitive to the presence of melt (increasing fluid content increases electrical conductivity EC and seismic wave attenuation, and decreases seismic velocities) but they relate to chemistry, temperature, and connectivity differently, thus complementing each other. In the field, these properties are probed using electromagnetic and seismic methods, offering a unique way to map melt distribution in the upper mantle in real time and understand its dynamics.

A compilation of geophysical results in 18 locations where melt is expected suggests different relationships between electrical and elastic parameters. In volatile-depleted contexts (mid ocean ridge, hotspot), a correlation between EC and S- and P-waves velocity can be interpreted in terms of melt fraction when combined to the laboratory database of the EC of melts as well as petrology modeling [1]. For example, geophysical results in Hawaii and the Afar Ridge are consistent with the presence of ~ 1vol % and 1-8vol % connected melt phase, respectively. In volatile-rich contexts (subduction), the lack of laboratory studies on relevant volatile-rich compositions makes the elaboration of a petrology-based model difficult to interpret field data. However, a comparison of electromagnetic and seismic data for several back-arc and trench-close fluid reservoirs in the mantle wedge shows a correlation between EC and the inverse quality factor for S and P waves, Qs-1 and Qp-1, respectively

EC = 1.5e-3×Qp-1 – 0.104 and EC = 1.2e-3×Qs-1 – 0.059

(Respective correlation coefficient of 0.78 and 0.96). A possible relationship between geophysical properties of reservoirs and slab age is also suggested. These field-based relationships offer promising constraints on melt distribution, but further input from laboratory studies is needed.

[1] Pommier A. and E. J. Garnero (2014) Petrology-based modeling of mantle melt electrical conductivity and joint interpretation of electromagnetic and seismic results, J. Geophys. Res. 119 4001–4016.