Results of previous experimental assessment of soil erodibility, using the methodology of washing out model samples by a water stream at different installations (hydraulic flume, jetting unit, and water tube), were usually strongly variable. To solve this problem, a new approach to the formation of samples was elaborated. It differs from the previous one by loading the sample through a removable side wall of the horizontally placed cartridge. The layered loading and compaction of samples remains unchanged. The normal position of sample layers to the flow contrary to the parallel position practiced formerly enabled us to obtain a more uniform sample resistance to the flow. The use of a Poseidon measuring unit developed to define mean depth of the wave flow provided not only many-fold increase in the number of measurements of washout intensity during the experiment (from 1 to 14–15), but also enabled us to assess the quality of sample preparation, which increased accuracy and reliability of the study. The percentage standard deviation for erodibility was almost always lower in case of the horizontal loading as compared to the vertical one both between and within the series (by 13 and 12

The erodibility of model soils has been studied using the methodology of washing out soil samples with a water flow in a hydrodynamic tube. Compared to the traditional method of vertical loading of cartridges, the new method of forming soil samples using horizontal loading has made it possible to obtain a more homogeneous resistance of samples during their erosion with the water flow; the layers of soil aggregates were formed perpendicularly to the water flow. The use of the Poseidon measuring device has made it possible to obtain erosion rates in real time, evaluate the quality of sample preparation, and analyze the data for their rejection, in particular, at the beginning and end of the experiments. In addition, the use of the device has allowed us to skip some operations needed to clarify the data of the experiments at their early termination (the consideration of the remainder of a sample, which requires its drying to determine the amount of eroded soil by the mass difference). The data files generated by the device at the end of the experiments significantly simplified their processing and considerably reduced the required time. The coefficient of variation in the erosion rate decreased by an average of 12

The effect of different factors and preparation conditions of monofraction samples from the arable horizon of leached chernozem on soil erodibility and its relationship with soil tensile strength (STS) has been studied. The exposure of samples at 38°C reduces their erodibility by two orders of magnitude. The drying of samples, on the contrary, increases their erodibility. It has been shown that erodibility decreases during the experiment. It has been found that the inoculation of soil with yeast cultures ( Naganishia albida , Lipomyces tetrasporus ) reliably increases the STS value in 1.5–1.9 times. The sterile soil is eroded more intensively than the unsterile soil: at 4.9 and 0.3 g/(m 2 s), respectively. The drying of soil followed by wetting to the initial water content (30%) has no significant effect on the STS value in almost all experimental treatments.

It has been shown in experiments in a hydraulic flume with a knee-shaped bend that the rate of soil erosion more than doubles at the flow impact angles to the channel side from 0° to 50°. At higher channel bends, the experiment could not be performed because of backwater. Results of erosion by water stream approaching the sample surface at angles between 2° and 90° are reported. It has been found that the maximum erosion rate is observed at flow impact angles of about 45°, and the minimum rate at 90°. The minimum soil erosion rate is five times lower than the maximum erosion rate. This is due to the difference in the rate of free water penetration into the upper soil layer, and the impact of the hydrodynamic pressure, which is maximum at the impact angle of 90°. The penetration of water into the interaggregate space results in the breaking of bonds between aggregates, which is the main condition for the capture of particles by the flow.

It has been experimentally shown with monoaggregate model samples of chernozemic soil as an example that the weakening of cohesion between soil particles is due to the molecular interaction of soil particles with water as a dipole substance rather than to the hydraulic forces of the flow. Therefore, soil erosion should be considered as a two-stage process. First, the bonds between particles are weakened due to the interaction of soil particles with water; then, the particles that lost bonds with neighboring ones are entrapped by the flow. Thus, the erosion rate of a consolidated soil is determined by the destruction of bonds between particles during their interaction with water, rather than the flow velocity, although this factor also affects the erosion rate.

It is known that tensile strength of soil samples is by three orders of magnitude greater than the shear stresses on the bottom of slope streams responsible for the detachment and transport of soil particles by water current. C.E. Mirtskhulava believed that detachment of soil particles by water current occurs due to the fatigue destruction of bonds between soil particles.Taking this fact into account, tensile strength is lower by two orders of magnitude. M.A. Nearing had an opinion that detachment of soil particles occurs in the points of separation of vortices from the bottom of the stream, where the shear stress is by two orders of magnitude higher than the average. These approaches did not explain overcoming by slope streams of the cohesion forces between soil particles. Studies of the influence of water temperature on the washout rate of model samples have shown that the soil erosion is highly dependent on the water temperature, which is close to the Van’t Hoff’s rule. This means that destruction of bonds between soil particles is probably the result of interaction between the soil solid phase and water molecules. Experiments have also shown that destruction of bonds between soil particles in the sample of chernozem monoaggregate soil occurs under a layer of still water. Upon the start of the water flow, particles that lost bond with the rest of the soil body immediately break away. The number of particles with disrupted bonds grows with the duration of the sample exposure to still water, although with some flattening. Experiments confirm the validity of the hypothesis of non-hydraulic nature of forces that disrupt inter-aggregate bonds during water erosion.

The article is devoted to the theoretical analysis and experimental investigation of the bottom and lateral erosion in shallow flows on slopes. The analysis of the ratio between the forces detaching and retaining a soil particle has showed that the erosion of the bed sidewall exceeds manifold the erosion of the bottom at the flow velocity close to the threshold value. When the flow velocity increases, the differences in the rate of erosion between the bottom and the sidewalls of the rill are leveled. The rate of sidewall erosion strongly depends on the slope of rill sides. The experimental studies of the effect of the sample surface inclination have completely confirmed the theoretical conclusions. It should be kept in mind that the lateral erosion under natural conditions is also limited by the laws of hydraulics. When the rill bed is widened due to the lateral erosion, the flow width increases and, hence, its velocity decreases to below the threshold value, which stops the erosion of the bed.

The close almost functional relationship of the erosion rate and, hence, the erodibility of model soil samples with the temperature of the water used in the experiments has been shown. This suggests that the rupture of bonds between the particles of eroded soil samples is due to the electrostatic forces appearing between the monomolecular water layers around the adjacent soil aggregates similarly oriented with respect to the soil solid phase rather than to the hydraulic forces. The erosion parameters of the samples also strongly depend on the soil moisture. The lowest erosion rate of the heavy loamy chernozem samples is observed at an initial water content of 22–24%. The erosion rate increases and the variability of the results is reduced with both decreasing and increasing the initial water content.

Результаты исследований показали очень тесную почти функциональную связь интенсивности размыва образцов и соответственно эродируемости модельных образцов почвы с температурой воды, использованной в экспериментах. Это дает основание полагать, что нарушение связей между частицами размываемых образцов почвы происходит за счет не гидравлических, а электростатических сил, возникающих между одинаково ориентированными относительно твердой фазы почвы одномолекулярными слоями воды вокруг смежных почвенных агрегатов. Эрозионные параметры образцов также сильно зависят от влажности почвы. Наименьшая интенсивность размыва образцов из тяжелосуглинистого чернозема наблюдается при исходной влажности 22–24%. Как при уменьшении, так и при увеличении исходной влажности скорость размыва увеличивается, а вариабельность результатов уменьшается.

Soil erodibility as a function of the soil density was studied in a hydraulic flume. The experiments were performed with mixtures of aggregates and different-sized chernozemic soil particles with the predominance of the fraction of 0.5–2 mm, which corresponded to their proportions in the plow horizon. It was found that, when the soil density increased, the erodibility abruptly decreased and was well described by a power function with an exponent of −9.25 (R 2 = 0.997). The approximate calculation of the contact area between the aggregates showed that the cohesion force between them is not a linear function of the total area of the interaggregate contacts.

The rates of the soil loss were studied in a wide range of flow velocities from 0.41 to 1.64 m/s. The results confirmed the main implications of the hydrophysical model: (1) the rate of the soil loss in the region of flow velocities exceeding the threshold value by 1.2–1.5 times and more is proportional to the cubic flow velocity; (2) the detachment of particles in the pre-threshold region is probable in nature, and the relationship between the soil loss and flow velocity is S-shaped. On this basis, a procedure was developed for the experimental determination of the soil erosion parameters: the erodibility, the threshold flow velocity, and the coefficient characterizing the variance of the particles’ resistance to detachment.

Field studies and model experiments, as well as theoretical considerations, suggest that bed sediments represented by soil aggregates in overland flows on slopes exert a considerable influence on the intensity of erosion processes. In this context, one of the key problems in the development of adequate erosion models is the problem of the rate of destruction of such aggregates in the flows. The results of experimental studies of the destruction of aggregates of chernozemic soils are analyzed. It is found that the destruction of soil aggregates in the flow proceeds in two stages. During the first stage, the aggregates are rapidly broken apart into smaller fragments. During the second stage, these fragments are subjected to abrasion. An equation describing the destruction of aggregates upon their movement in the flow in dependence on the aggregate size and the distance of aggregate transport is suggested. The effect of some groups of soil microorganisms on the aggregate resistance to the destruction is shown.