Ages of Faulting and the Nature of Fluid-Rock Interactions in the Deep Nankai Accretionary Prism, IODP Site C0002, Japan

The NanTroSEIZE project investigated factors controlling seismicity in faults within the accretionary prism. Samples from Expeditions 348 and 358 provide evidence for widespread fracturing and fluid-modulated chemical and mineralogical changes in faults. Geochemical data from secondary carbonates provide constraints on the timing of deformation, fluid flow, and changes in material properties to depths to 3262.5 mbsf.

Results from Exp 348 showed steeply dipping bedding, common fracturing and fault-related textures, and an abundance of carbonate veins (Sample et al., 2018). Oxygen isotope signatures suggest multiple phases of veins are present. Veins with only small depletions in ¹⁸O (d¹⁸O values > -5 ‰) may have formed early in the burial history; more depleted veins are consistent with formation at in situ temperatures after deep burial. Carbonate clumped isotope data have a temperature range of 50°C to 110°C, a temperature gradient compatible with clay mineralogy (Underwood, 2018) and with a published model based on other information (Sugihara et al., 2014). Because the entirety of Hole C0002P was open during drilling, it is unclear whether the abundant carbonate veins found in cuttings were mostly cavings from an upper, vein-rich horizon or were distributed at many levels in the wall rock. Clumped isotope data support the latter inference. U-Th data show seven veins formed between 561 ka and 108 ka before present, thus postdating filling of the Kumano Basin. Therefore the veins are likely to accurately reflect recent in situ temperatures. Waters during carbonate formation were slightly enriched in ¹⁸O.

Drilling during IODP Exp 358 recovered carbonate vein samples in cuttings from depths up to 3262.5 mbsf. Because the cuttings were from a new hole sidetracked from C0002P, caving from Hole C0002P is not possible, and thus the depth interval of veining in Hole C0002Q is relatively narrowly constrained. We will show results of isotope analysis on carbonates and infer the timing and origin of fluids forming fracture-filling veins. While the plate-boundary fault was not reached during Exp 358, our data suggest lithologic hardening of clay-rich horizons may result in part from extensive veining. It is likely that silica diagenesis and quartz veining is more prevalent between 3000 mbsf and the plate-boundary fault.