Grading patterns of river flood deposits in a subaqueous delta environment varies with distance from the mouth: example from Lake Shinji, Japan, as a natural laboratory

Hyperpycnal flows have been recognized as an important sediment delivery process in marine environment. In order to clarify whether the momentum of river flows during floods propagates uniformly to offshore or not, we acquired three geo-slicer cores along a longitudinal profile on the subaqueous portion of the Hii River delta built since the 1630s in Lake Shinji, western Japan. Because the hydrologic energy of the lake is significantly low, deposits derived from the Hii River floods were well preserved on its delta front slope region capped by mud. Grading patterns of 26 individual sand beds in the cores vary with water depth. Triple stacks of inverse-to-normal grading is seen in beds of shallower horizons than 5 m below the water surface. Single inverse-to-normal grading mainly appears between 4 and 5 m depth, and normal grading dominates between 5 and 6 m depth. Assuming that flood hydrographs for the Hii River have not changed since the 17th century, this variation suggests the non-uniform propagation of the momentum of the river flow to its outflow. Inverse and normal grading is interpreted to reflect the waxing and waning of the parent flow, respectively. Thus, the hydrograph of the flood outflow is suggested to become simple with distance from the mouth. Triple stacks of inverse-to-normal grading in shallower horizons can be interpreted as consequences of movement of the plunge point of flood plumes during the course of flood events. Spatially decelerating sediment-laden river plumes steeply increase their velocity after they plunge beneath the water surface (Lamb et al., 2010). In depth-limited proximal areas of a subaqueous delta, back-and-forth translation of the plunge point over a fixed point due to the waxing and waning of river discharge leads to three cycles of waxing and waning of flow velocity. In the distal parts of the delta, where the plunge point does not reach, velocity of plunged hyperpycnal flow increases and then decreases reflecting directly the waxing and waning of river discharge, and would have formed single inverse-to-normal grading. There are two possible explanations for overall normal grading seen in further deeper horizons, 1) only a part of hyperpycnal flow generated around the flood peak could have reached the area as a surge-like flow because of the lateral spreading of the flow, or 2) the slower initial part of the hyperpycnal flow had been overtaken by the succeeding faster part to yield a monotonically waning flow at a fixed point.