Inferring Earthquake Physics from Deep Drilling Projects of Active Faults

Wednesday, 17 December 2014
Giulio Di Toro1, Steven A.F. Smith2, Li-Wei Kuo3, Silvia Mittempergher4, Francesca Remitti4, Elena Spagnuolo5, Thomas M Mitchell6, Alessandro Gualtieri4, Jafar Hadizadeh7 and Brett M Carpenter8, (1)University of Padua, Padua, Italy, (2)University of Otago, Dunedin, New Zealand, (3)National Taiwan University, Taipei, Taiwan, (4)University of Modena and Reggio Emilia, Modena, Italy, (5)National Institute of Geophysics and Volcanology, Rome, Italy, (6)University College London, London, United Kingdom, (7)University of Louisville, Louisville, KY, United States, (8)INGV National Institute of Geophysics and Volcanology, Rome, Italy
Deep drilling projects of active faults offer the opportunity to correlate physical and chemical processes identified in core samples with experiments reproducing the seismic cycle in the laboratory and with high‐resolution seismological and geophysical data.

Here we discuss the constraints about earthquakes source processes at depth gained by fault cores retrieved from the deep drilling projects SAFOD (2.7 km depth, San Andreas Fault), J-FAST (0.9 km depth, following the Mw 9.0 Tohoku 2011 earthquake), TCDP (1.1 km depth, following the Mw 7.6 Chi-Chi 1999 earthquake) and WFSD (1.2 km depth, following the Mw 7.9 Wenchuan 2008 earthquake). Recovered samples were tested at room temperature with the rotary shear apparatus SHIVA installed in Rome (INGV, Italy). All the tested samples were made by clay-rich gouges (usually including smectite/illite), though their bulk mineralogy and modal composition were different (e.g., SAFOD samples included saponite, WFSD carbonaceous materials). The gouges were investigated before and after the experiments with scanning and transmission electron microscopy, X-Ray diffraction, micro-Raman spectroscopy, etc.

A common behavior of all the tested gouges was that their friction coefficient was low (often less than 0.1) under room-humidity and wet conditions when sheared at slip rates of ca. 1 m/s (seismic deformation conditions). Moreover, when the natural fault rocks next to the principal slipping zones were sheared from sub-seismic (few micrometers/s) to seismic slip rates, the experimental products had similar microstructures to those found in the principal slipping zones of the drilled faults. This included the formation of mirror-like surfaces, graphite-rich materials, foliated gouges, nanograins, amorphous materials, etc. In most cases the mechanical data were consistent with several seismological (> 50 m of seismic slip for the fault zone drilled by J-FAST) and geophysical observations (absence of a thermal anomaly in the fault cores of J-FAST and WFSD) which were attributed to an extremely low coseismic fault strength. However, the deformation mechanisms responsible for the measured weakening in the experiments and for the production of the microstructures similar to those found in drilled seismic faults have not been deciphered yet.