Response of loess materials to simulated translocation by water: Micromorphological observations

Abstract

AbstractStructural differences between loess material deposits produced by splash, flow without splash and by rainwash (i.e. combined splash and flow) respectively were reported earlier (Mücher and de Ploey (1977)). Deposits from splash alone proved to be non‐layered and had incurred almost no additional particle sorting. Deposits from flow without splash were well layered, while individual laminae displayed excellent sorting. Rainwash produced poorly layered deposits with poorly sorted individual laminae. Only horizontal and subhorizontal bedding, and no cross bedding, were observed. Similar features appeared in samples from naturally layered loess deposits collected from excavations.The present paper focuses on the granulometric character of the artificially created layers. Experiments normally involved the combined use of an erosion and a sedimentation flume. Slope angles of the flumes varied mainly form 0.25 to 4 degrees. Precipitation intensities varied from 7 to 120 mm per hour. Unit discharges ranged from 0.2 to 9.2 cm3/cm,s. Flow was laminar or smooth turbulent, either subcritical or supercritical (Froude numbers between 0.5 and 3.8). Raindrop impact had little influence on the flow velocity and thickness of the surficial water film, but it significantly increased transportation capacity and flow competence (Savat (1976, 1977)).Undisturbed samples were collected during the experiments, and impregnated with unsaturated polyester resin. Thin sections of 20 μm thickness were studied under a polarizing microscope. Comparative observations were made on naturally layered loess materials.During rainwash, clay size particles are removed and exported, whereas laminae with particles from 10 to 50 microns are more frequently produced than those with particles from 10 to 30 microns.Flow without splash produced relatively more laminae with grains smaller than 30 microns diameter.Striking in all samples is the absence of laminae composed of 30–50 μm grains. Aggregates were only deposited by smooth turbulent flow, either disturbed or undisturbed by raindrop impact.In view of the experimental results, the naturally formed loess deposits may be interpreted as follows: the bulk of the loess from the excavation near Tongrinne, Belgium, was reworked under influence of turbulent rainwash. Only one sample contained laminae consisting of clay, particles smaller than 10 μm, and of particles from 10–30 μm respectively, suggesting deposition from flow without splash. This was taken from a gully infilling.A gully fill from the excavation Heverlee near Leuven, Belgium, is characterized by poorly sorted laminae, dominated by the coarser silt fractions. This suggests infilling under influence of turbulent rainwash.