Modeling the impact of melt on seismic properties during mountain building

• Published in: Geochemistry, Geophysics, Geosystems
• Main Authors: Lee, AL, Walker, AM, Lloyd, GE, Torvela, T
• Format: Article in Journal/Newspaper
• Language: English
• Published: American Geophysical Union (AGU) 2017
• Subjects: Norway; Seiland; Northern Norway
• Online Access: https://eprints.whiterose.ac.uk/112722/; https://eprints.whiterose.ac.uk/112722/7/Lee_et_al-2017-Geochemistry,_Geophysics,_Geosystems.pdf; https://doi.org/10.1002/2016GC006705

Initiation of partial melting in the mid/lower crust causes a decrease in P-wave and S-wave velocities; recent studies imply that the relationship between these velocities and melt is not simple. We have developed a modelling approach to assess the combined impact of various melt and solid phase properties on seismic velocities and anisotropy. The modelling is based on crystallographic preferred orientation (CPO) data measured from migmatite samples, allowing quantification of the variation of seismic velocities with varying melt volumes, shapes, orientations, and matrix anisotropy. The results show non-linear behaviour of seismic properties as a result of the interaction of all of these physical properties, which in turn depend on lithology, stress regime, strain rate, pre-existing rock fabrics, and pressure-temperature conditions. This non-linear behaviour is evident when applied to a suite of samples from a traverse across a migmatitic shear zone in the Seiland Igneous Province, Northern Norway. Critically, changes in solid phase composition and CPO, and melt shape and orientation with respect to the wave propagation direction can result in huge variations in the same seismic property even if the melt fraction remains the same. A comparison with surface wave interpretations from tectonically active regions highlights the issues in current models used to predict melt percentages or partially molten regions. Interpretation of seismic data to infer melt percentages or extent of melting should, therefore, always be underpinned by robust modelling of the underlying geological parameters combined with examination of multiple seismic properties in order to reduce uncertainty of the interpretation.