Shear-wave attenuation anisotropy: a new constraint on mantle melt near the Main Ethiopian Rift

Fluids hosted in fractures, or low aspect ratio inclusions, exist in many different settings within the Earth. In the near surface, understanding systems of fluid-filled fractures is important to various industrial applications such as geothermal energy production, monitoring CO₂ storage sites and exploring for metalliferous sub-volcanic brines (e.g., Blundy et al., 2021). In the mantle, melting is an important geodynamic process, exerting control over its composition and dynamic processes. Upper mantle melting weakens the lithosphere, facilitating rifting (e.g., Kendall et al., 2005) and other surface expressions of tectonic processes. In the lowermost mantle, it has been suggested that ultra-low velocity zones could contain partial melt. A challenge, however, in all these settings is finding a geophysical observation which is sensitive to the presence of fluids and the host fracture networks.

The presence of fluids has a significant effect on the overall elasticity of the medium. It is well known that aligned fluid-filled fractures, or inclusions with small aspect ratios, produce seismic velocity anisotropy, even for very low volume fractions (e.g., Hudson, 1982, Chapman 2003). This mechanism is often used by shear-wave splitting studies to interpret the orientation of maximum horizontal stress within the crust (e.g., Crampin 1999, Asplet et al., 2024). The same rock physics models, however, also predict attenuation anisotropy that is frequency-dependent and sensitive to important fracture properties, such as fracture length and orientation. Therefore, if attenuation anisotropy can be measured, it offers an exciting new avenue to seismically detect fluids in the subsurface.

Here we show that attenuation anisotropy can be measured in conjunction with shear-wave splitting analysis. Using an instantaneous frequency matching method (after Mathenay and Nowack, 1995) the differential attenuation between fast and slow shear-waves can be measured. We explore the potential of this technique using synthetic data and SKS data collected at FURI, Ethiopia. We also demonstrate the potential systematic error, in both fast polarisation and delay times, that attenuation anisotropy can have on shear-wave splitting measurements and outline an approach for correcting measurements. For SKS data recorded at FURI shear-wave splitting and attenuation anisotropy is measured that requires poroelastic squirt flow of aligned melt inclusions oriented perpendicular to the Main Ethiopia Rift. This is a result which would not be interpreted by only considering SKS shear-wave splitting. These intriguing results highlight the potential for attenuation anisotropy as a tool to detect and characterise fluids in the subsurface.