[S-CG45] Science of slow-to-fast earthquakes

[SCG45-35] Shear heating recognized in subduction metamorphism and temperature at the base of the seismogenic zone

*Simon Richard Wallis¹, Yukinojo Koyama¹, Kazuhiko Ishii (1.The University of Tokyo)

Subduction type metamorphism is generally warmer than the conditions predicted by thermal models. One explanation for this discrepany is the influence of shear heating. However, clear evidence for shear heating has been lacking and preferential exhumation of warm rocks along subduction zones is a more widely accepted explanation. Previous studies have focused on peak metamorphic conditions. Our thermal modeling shows that strongly curved prograde subduction-related paths are a much better indicator of shear heating. The Cretaceous Sanbagawa belt of SW Japan is a well studied geological example with good constraints on the age of the subducting plate and rates of subduction, all of which make it an excellent area to examine the use of thermal modelling combined with metamorphic studies to quatify shear stresses and shear heating. Comparison of our model results with P-T paths reveals a good match for shear heating associated with an effetive coefficient of friction of ~0.13. The predicted temperatures at depths of around 30 km are 250-300°C higher than expected for models without frictional heating. The range of P-T paths recorded is compatible with a subduction channel of about 1 km thickness. One of the limitations and weaknesses of our modeling is that it requires assumtions to be made about the rheology along the plate interface. Appropriate rheological models are not well known. We used Byerlee's Law for the shallow frictional part. This domain is associated with most of the shear heating. The heating is limited by the down dip limit of the seismogenic zone, which in our models, like most others, is controlled by the onset of ductile deformation of quartz. An alternative appoach is to esitmate the depth of the seismogenic zone from modern subduciton zones and use that to constrain the domain of shear heating. The best analogue for the Sanbagawa belt is the Tonga subduction zone, which has a similar very fast rate (~20 cm/yr) of subduction of an old plate. In the Tonga subduciton zone the base of the seismogenic zone is ~30 km. If we use this value in our model instead of assumtions about rheology, we obtain a good match with the observed data for an effective coefficient of friction of 0.06. Estimates of shear stresses from microstructural studies of the Sanbagawa belt agree well with this value. If this is correct the temperature at the base of the paleo seismogenic zone was and likely is much higher than generally thought. This challenges assumptions about where to look for geological evidence of slow earthquakes.