Sediment transport and deposition of various textured soils in shallow flow

Soils with a wide range of textural composition were used in a series of laboratory experiments conducted to study the transport and deposition of sediment eroded by simulated rainfall and deposited along a concave 0.5 x 3.0 m bed. A procedure was developed to determine the density of soil aggregates for the silt loam and silty clay soils. Samples of sediment entering, leaving and deposited on the concave bed were collected. Change in characteristics of sediment leaving the depositional area with time compared with the inflow sediment were studied. Deposition of sediment on the concave bed, its evolution of a deposition profile, size distribution of deposited sediment along the bed, and distribution between primary particles and aggregates were examined. Findings for these shallow flow conditions were at variance with the classic sediment transport theory. In this study, for silty clay and fine soils, median size of deposited sediment increased in the downslope direction, while for the silt loam the median size decreased downslope. For silt loam and silty clay soils, median size of sediment leaving the depositional area increased with time, while for the fine sand soil the median size decreased with time. The degree of exposure of particles to flow and raindrop impacting the area were found to be the major cause for these results. Dimensionless sediment transport relationships were developed for uniform and non-uniform size of non-cohesive sand sediment and for non-uniform cohesive agricultural soils. For the uniform sand and non-uniform sandy soils, a typical bed load equation with consideration of excess tractive force fit the data reasonably well. This may also imply that rolling, sliding and hopping were the major modes of motion for the sandy soil particles under shallow flow conditions. For the silt loam soil, a power function of dimensionless shear stress with an exponent of about 1.5 gave good prediction of sediment transport capacity. For silt loam and sandy soils the presence of rainfall on the depositional area did not change considerably the coefficient and the exponents which were obtained when data with and without rainfall were considered. For silty clay soils a reliable relationship was not found to describe the sediment transport capacity. A simplified deposition model was developed to simulate the evolution of bed profile for the sandy soil.