Parnell Grits—Large Subaqueous Volcaniclastic Gravity Flows with Multiple Particle-Support Mechanisms

Abstract Parnell Grits are volcaniclastic beds 2 to 20 m thick interbedded comformably with bathyal flysch of the Waitemata interarc basin (early Miocene) of northern New Zealand. They comprise lava clasts, crystals, rip-up clasts, shallow-marine fossils, and rare extrabasinal clasts in a matrix of sand and silt. Inverse-to-normal size grading is always present. Other primary structures are generally absent. Rarely, the fine-grained upper levels show horizontal and micro-cross-lamination. Evidence of high post-depositional internal pore pressures consists of intrusions of grit downward into underlying sediments, inward into most rip-up clasts, and upward into upper levels of the Grit and rarely into overlying sediments. Rip-up clasts range up to 90 m long. They comprise more than half of the volume of the deposit in some places, occur at all levels in the flow, and may be folded. Lava clasts up to 20 m long were carried to the center of the basin (>40 km), implying a bulk strength for some moving flows of >1.25 kg/cm2. After deposition, the two largest known lava clasts indented underlying beds, implying a bulk strength after deposition of between 0.5 and 0.78 kg/cm2. The lower strength after deposition is attributed to decay of grain interaction. We interpret Parnell Grits as the deposits of volcaniclastic gravity flows which began as debris avalanches resulting from sector collapses on subaerial slopes of the Waitakere Arc to the west of the Waitemata Basin. The volcanic-debris flows crossed the strand zone, picked up a shallow-marine biota and flowed on to bathyal depths. On entering the sea, there was a change in flow character resulting from the uptake of water. The reduction in matrix strength led to the development of a basal traction carpet of clasts in grain-flow and normal-size grading of dense clasts. Rip-up clasts were acquired by injection of grit into the substrate. There is very little evidence for turbulence in the flows and no evidence of flow transformation. Deposition was by frictional freezing in the traction carpet, cohesive freezing in the bulk of the deposit, and in some beds fluidal settling of the fine elutriated tail. Decay of some clast-support mechanisms after deposition reduced the bulk strength of the deposit. There were probably more than 100 such flows, all with similar properties.