Sediment transport pathways on the modern microtidal sand flat: comparison between grain-size trend model and tracer experiment

Sediment transport pathways on the modern microtidal sand flat along the Kushida River Delta were estimated by a statistical model proposed by Yamashita et al. (2010) (P-GSTA method), which is based on the spatial distribution of sediment grain-size. The result was verified by the tracer experiments. The Kushida River flows into Ise Bay, and forms a bayhead delta. The tidal range is about 2 m during spring tide (microtidal). The sand flat is spread ca. 0.4 km2 on the right bank of the river. The sand flat is mainly composed of medium- to coarse-grained sand. Based on the facies analysis of subsurface sediments, it was interpreted that sediment transport is dominated by fluvial and wave activities. In the P-GSTA method, a linear function in which six parameters of grain-size distribution (mean, coefficient of variance, skewness, kurtosis, and mud and gravel logratios) are summated with different weighting factors was used to infer sediment transport direction. For automated determination of the weighting factor of each grain-size parameter, the principal component analysis (PCA) of grain-size parameters was conducted. The weight of each parameter depends on its variance. It was revealed that the first principal component (PC1) account for the spatial variation of the grain-size distribution as a result of sediment transport. The factor loading of PC1 indicates that the grain-size distribution of sediments on the surface of the microtidal sand flat becomes finer, better sorted, less gravelly downcurrent through the sediment-transport processes by fluvial and wave activities. Then, the eigenvector of PC1 was employed as weighting factors of grain-size parameters to calculate linear function of grain size parameters representing sediment transport. The outline of the sediment transport pathways reconstructed by this method is as follows: (1) sediments are mainly supplied from the river mouth, then, (2) they are drifted from northwest to southwest, and finally (3) dispersed northeastward. For validation of the result of the P-GSTA method, tracer sediments (blue-colored sand) were situated on 6 localities in the tidal sand flat at May 18th, 2011 (neap spring tide). 2 weeks after, bottom surface sediments were collected around ca. 10 m from the source points of tracer sediments. Then, the number of colored sand grains mixed in natural sands was counted as a proxy of sediment transport. The result of tracer experiment well validated the trends (1) and (3) indicated by the P-GSTA method, but it conflicts to the trend (2) at the middle part of the sand flat. This is probably because the result of P-GSTA method reflects sediment transport patterns in the longer timescale, in which episodic sedimentary events such as storms or fluvial floods are included. Indeed, the trend (2) inferred by P-GSTA method coincides to the direction of ephemeral braided channels on the sand flat that was observed after fluvial flooding event in July 2011. These results indicate that the result of the P-GSTA method reflects both the averaged long-term sediment transport patterns and episodic depositional events.