Rill Development And Its Morphological Characteristics At Loess Hillslope

Rill development results in severe soil loss at hillslopes. Rill morphology formed by rill development has important influence on flow hydraulic and hydrodynamic characteristics, which inversely affects the hillslope soil erosion process. Therefore, studies on rill development and its morphological characteristics are of great importance in revealing the hillslope soil erosion mechanism, and in establishing the soil erosion prediction model. A laboratory rainfall simulation and field observation were conducted in this study. The development process of rill network at the loessial hillslope was investigated. The rill morphological characteristics were quantified. The relationship between rill erosion and its morphological characteristics was analyzed. A rill erosion prediction equation based on rill morphological indicators was established and validated. The effects of rainfall(raindrop impact and rainfall intensity) and topographic factors(slope gradient and slope length) on rill erosion and its morphological characteristics were exposited. Based on the important contribution of raindrop impact to rill erosion, the soil erosion continuity equation in Water Erosion Prediction Project was revised. Additionally, a hillslope soil erosion model was established and validated. The main results were as follows:(1) The development process of rill network at the loessial hillslope was studied. Taking a waterfall as a symbol of rill occurrence, changes of waterfall width and depth were analyzed. Most waterfall widths were between 2~8 cm, and the average value was 5.0 cm. Most waterfall depths were between 2~6 cm, and the average value was 4.4 cm. Based on indicators such as rill distribution density, distance between rills, rill bifurcation, and combination numbers, the development characteristics of rill network were quantified. Furthermore, the differences in development characteristics of rill network for hillslopes of rainfall simulation and field observation were compared. The average values of the four indicators were 4.0 numbers m–2, 38.6 cm, 1.5 and 1.1 numbers m–2, respectively. With rill network development, rill distribution density generally first increased and then decreased, distance between rills gradually decreased, and rill bifurcation and combination numbers increased. With an increase in slope length, rill distribution density, rill bifurcation, and combination numbers generally first increased and then decreased; however, distance between rills first decreased and then increased. The development characteristics of rill network between hillslopes for rainfall simulation and field observation were similar.(2) The rill morphological characteristics at the loessial hillslope were quantified. Rill inclination angle, rill density, degree of rill dissection and rill tortuosity complexity were chosen to investigate the temporal and spatial variation of rill morphological characteristics and propose the optimal indicator. The four rill morphological indicators generally increased with the development of rill network. Most rill inclination angles concentrated from 15°~25°, and values of rill tortuosity complexity were < 1.5. Furthermore, the increment rate in degree of rill dissection was greater than that in rill density. With an increase in slope length, rill inclination angle generally decreased with fluctuations, rill density and degree of rill dissection first increased and then decreased, rill tortuosity complexity did not exhibit an obvious change trend. Degree of rill dissection was the best derivative morphological indicator to evaluate rill development. The rill erosion prediction equation covered degree of rill dissection was established and validated, and the performance of this model was satisfactory.(3) The effects of raindrop impact on rill erosion and its morphological characteristics were analyzed. Eliminating raindrop impact through placing the nylon net over the soil pan caused rill erosion and soil loss to decrease by 20.2%~38.6% and 28.1%~47.7%, respectively; however, induced an increase in the contribution of rill erosion to hillslope soil erosion. Furthermore, the effect of raindrop impact on rill erosion increased as rainfall intensity increased. Eliminating raindrop impact also indirectly influenced rill morphology. That is, the degree of variation in rill width and rill depth decreased, and rill channel shape was more inerratic. Additionally, rill density, degree of rill dissection, rill inclination angle and rill width-depth ratio decreased. The results indicated that eliminating raindrop impact resulted in the degree of the hillslope dissection, rill erosion intensity, and the ductility of rills in the horizontal direction, etc. decreasing.(4) The effects of raindrop impact on rill flow hydraulic characteristics and dynamic mechanisms of rill erosion were explored. Eliminating raindrop impact through placing the nylon net over the soil pan induced rill flow velocity, Reynolds number and Froude number and its morphological characteristics at the loessial hillslope were sensitive to rill flow velocity and stream power.(5) The soil erosion continuity equation in Water Erosion Prediction Project was revised, and a soil erosion prediction model at the loessial hillslope was established and validated. Since the important contribution of raindrop impact to rill erosion, the rill erosion segment of the soil erosion continuity equation in Water Erosion Prediction Project was revised. That is, when detachment was the main pattern in the process of rill development, the rainfall intensity indicator was added in the rill erosion segment, which reflected the effects of rainfall on rill erosion. The Nash-Sutcliffe simulation efficiency of the soil erosion prediction model established in this study to soil erosion rate was 0.89. This indicated that the model had better prediction accuracy.