Evaluation of in situ smectite dehydration as a pore water freshening mechanism in the Nankai Trough, offshore southwest Japan

Abstract

Pore water freshening has been observed within sediments near the trench at numerous subduction zones. Constraining the relative contributions of long‐distance updip flow of freshened fluids and in situ clay dehydration holds important implications for margin‐scale fluid flow but remains unresolved because the evolution of pore water chemistry expected due to in situ clay dehydration and the budget of fresh water release have generally not been carefully quantified. Here, we address this problem at the well‐studied Nankai Trough offshore SW Japan, where Deep Sea Drilling Project and Ocean Drilling Program drilling have sampled the sedimentary section along two transects: the Muroto transect and the Ashizuri transect, located ∼100 km southwest along strike. Broad low‐chloride anomalies with minimum concentrations significantly below typical seawater values were documented along the Muroto transect at Sites 1173, 1174, and 808 and indicate progressive freshening with increasing burial. In contrast, freshening at Site 1177 along the Ashizuri transect (∼24 km seaward of the trench) is characterized by a series of sharp excursions from seawater values. We use a one‐dimensional model to simulate sedimentation, thermal history, clay dehydration, and chemical diffusion, in order to quantify the component of pore water freshening attributable to in situ smectite dehydration. We find that simulated reaction progress is in close agreement with observations. The peak magnitude of pore water freshening is reasonably well predicted within the lowermost ∼150 m of the section but consistently underpredicted at intermediate depths. The residual pore water freshening profiles (not explained by smectite dehydration) reach peak magnitudes of ∼50–75 mM and approximately straddle the décollement at Sites 1174 and 808. We suggest that these residuals may reflect either an additional in situ dehydration process or a contribution from updip fluid migration along the fault. Our conclusion differs from that of previous studies, which have suggested that most of the observed geochemical anomaly could be generated by in situ smectite dehydration with as little as 15 wt% initial smectite. This substantial discrepancy is readily explained by the high sensitivity of pore water freshening to assumed porosity at the time of dehydration.