d18O, lithologies, mineralogy and determination of temperature and depth of formation of DSDP Site 69-504 siliceous rocks

Seven opal-CT-rich and five quartz-rich porcellanites and cherts from Site 504 have a range in oxygen-isotope values of 24.4 and 29.4 per mil. In opal-CT rocks, d18O becomes larger with sub-bottom depth and with age. Quartz-rich rocks do not show these trends. Boron, in general, increases with decreasing d18O for porcellanites and cherts considered together, supporting the conclusion that boron is incorporated within the quartz crystal structure during precipitation of the SiO2. Silicification of the chalks at Site 504 began 1 m.y. ago - that is, 5 m.y. after sedimentation commenced on the oceanic crust. Temperatures of chert formation determined from oxygen-isotope compositions reflect diagenetic temperatures rather than bottom-water temperatures, and are comparable to temperatures of formation determined by down-hole measurements. Opal-A in the chalks began conversion to opal-CT when a temperature of 50°C was reached in the sediment column. Conversion of opal-CT to quartz started at 55 °C. Silicification occurred over a stratigraphic thickness of about 10 meters when the temperature at the top of the 10 meters reached about 50°C. It took about 250,000 years to complete the silica transformation within each 10-meter interval of sediment at Site 504. Quartz formed over a stratigraphic range of at least 30 meters, at temperatures of about 54 to 60°C. The time and temperatures of silicification of Site 504 rocks are more like those at continental margins than those in deep-sea, open-ocean deposits.

Measurements of velocity-dependent and slip-dependent frictional strength in laboratory friction experiments on natural samples from IODP Expedition 316 and ODP Hole 190-1174B

Slowslip forms part of the spectrum of fault behaviour between stable creep and destructive earthquakes. Slow slip occurs near the boundaries of large earthquake rupture zones and may sometimes trigger fast earthquakes. It is thought to occur in faults comprised of rocks that strengthen under fast slip rates, preventing rupture as a normal earthquake, or on faults that have elevated pore-fluid pressures. However, the processes that control slow rupture and the relationship between slow and normal earthquakes are enigmatic. Here we use laboratory experiments to simulate faulting in natural rock samples taken from shallow parts of the Nankai subduction zone, Japan, where very low-frequency earthquakes - a form of slow slip - have been observed.We find that the fault rocks exhibit decreasing strength over millimetre-scale slip distances rather than weakening due to increasing velocity. However, the sizes of the slip nucleation patches in our laboratory simulations are similar to those expected for the very lowfrequency earthquakes observed in Nankai. We therefore suggest that this type of fault-weakening behaviour may generate slow earthquakes. Owing to the similarity between the expected behaviour of slow earthquakes based on our data, and that of normal earthquakes during nucleation, we suggest that some types of slow slip may represent prematurely arrested earthquakes.