A Method To Design Scale Model Experiment For Soil And Water Conservation In The Small Watershed Of Loess Plateau, China

Xu Xiangzhou,Zhang Hongwu

2005-01-01

Abstract:The key to a successful model study is the development of valid similarity requirements. This paper presents an approach to the scale model problem that is different from any previously attempted. An analysis and the interpretation of similarity requirements in a rainfall simulation system is presented, with controlled rainfall intensity and saturated bare land cover to model soil erosion in the small water-shed of Loess Plateau, China. Consequently experimental data on the model and the prototype are used to validate the similarity.In the small watershed of Loess Plateau, usually one run of short but very intense rainstorm could account for most of the total precipitation and soil loss in a year, hence the soil product from the small watershed for a whole year could be simulated by one run of artificial rainfall in laboratory. Soil and water erosion processes of the small watershed were simulated by the immature model with a bare and saturated land cover, while the relative erosion depth which is the ratio of erosion thickness to the length of the watered area, was consistent to that of the prototype, and the landform including the size of the check - dam, was proportional to that of the prototype one. The erosion proportion between the prototype and the model could be deduced according to the relative erosion depth. Each of the models was catena including elemental gemorphological units with characteristics of the typical small watershed in Loess Plateau, and plot 2 # was the normal miniature of plot I #. Homologous matrix loess was used as the erodible material of the two models, and consistent measurement equipments for rainfall, sediment product and runoff were employed in the two models in order to make sure the comparability of the data collected from the prototype and the model. Erosion was monitored by the use of a relief meter inside the plots and, at their outlets, by sampling the discharge. Two series of experiments were conducted in plot I # and plot 2 # respectively. Firstly, each 6 similar to 7 rainfalls with different intensities between 1.0 and 2.0 mm/min was applied in order to find rainfall erosivity of the antecedent land cover of the model and the prototype. Secondly, 10 rainfalls with the same intensity applied on the landscape of the model and the prototype respectively while the twelve dams were orderly constructed. The pre - established rainfall intensity of the model experiment could be calculated by the power function of rainfall intensity to the erosivity according to the converted model sediment product from that in the antecedent landscape. As a result the model soil product converted via the volume scale after each designated rainfall was close to that in the prototype experiment.