Quantitative Evaluation of Water Flux in the Franciscan Subduction Zone Based on Volume Change of Metamorphic Rocks

It is generally accepted that the development of slow earthquakes in subduction zone environments is closely associated with the development of high fluid pressure. It has also been proposed that the transport and deposition of silica in water-rich fluids may be important in controlling the time scales of deep slow earthquakes: Audet & Burgmann (2014) propose the recurrence period of deep slow earthquakes is controlled the amounts of silica accreted to the lower forearc crust with estimated amounts of 5－15 vol.%. Silica in the form of quartz is widely present in subducting rocks and has a high solubility compared to other common phases and so the amount of silica accretion in the low porosity rock domains expected in the lower crust should closely correspond to an equivalent volume increase. Therefore, estimating the volume increase in metamorphic rocks that have been exhumed from equivalent parts of the subduction system can be used to estimate silica precipitation in these domains. If the physical conditions associated with precipitation are known, the amount of precipitated silica can also be used to estimate the time-integrated water flux responsible for its transportation (Ferry & Dipple, 1991).

Application of the deformed vein sets method (Soejima & Wallis, 2022) to metagreywacke in Del Puerto Canyon in the Franciscan belt allows a quantitative estimate of volume change to be made. The Del Puerto metagreywacke represents a unit accreted to the hangingwall of the Franciscan accretionary complex and our results yield a volume increase of 7－21 vol.%, in good agreement with Audet & Burgmann (2014). The volume of quartz precipitated in the lower forearc crust estimated in this study can be related to the time-integrated flux of water moving upward from the deep subduction zone. Assuming that water percolated vertically in this region, the time-integrated fluid flux is estimated to be 1.0×106－1.9×107(m3/m2). This result is tens of times higher than the fluid fluxes estimated by combining results of thermal models of subduction zones, the distribution of hydrous minerals in subducting rocks and the stability of hydrous minerals (e.g. Peacock, 1990, van Keken et al., 2011). One possible explanation for this discrepancy is that subduction fluids are not transported vertically but channeled along the plate boundary.