摘要
根据多元正交设计,对马兰黄土进行了降雨(人工模拟)溅蚀试验。得到了计算降雨溅蚀分散量(Y)和降雨溅蚀向下坡搬运量(Z)的两个回归方程。方程揭示影响降雨击溅侵蚀诸因素之间的交互作用,以及各个因素在降雨溅蚀过程中的地位和对各种降雨溅蚀临界值的明显影响。
Laboratory investigation of rainsplash erosion on Malan Loess has been conducted based on a multivariate orthogonal design. Five factors including rainfall intensity (I) , cover density (V), rainfall duration (T) , antecedent soil moisture content (C) , and slope angle (A)were considered. In total, three replicates of 27 sets of experiments, each representating a different combination of the factors selected, were conducted.Regression equations using rainsplash detachment and rainsplash transport as dependent variables and derivatives of the five selected factors as independent variables were computed.What is significant is the inclusion of six and four interaction terms in the equation for splash detachmet Y and splash transport Z respectively.The analysis of results obtained from two equations shows that the interaction effects among some of the controlling factors of rainsplash erosion are of paramount importance. Apart from individual effects, both rainfall duration and rainfall intensity influence the role of slope angle in inducing splash detachment and splash transport. Throughout this paper, the existence of threshold slope angles has been shown. However, it is important to emphasize that for specific locations these slope angles can be approximately constant, they do vary in relation to the magnitude of rainfall events and vegetation cover condition. The protective effect of vegetation cover on splash transport is influenced by rainfall intensity primarily because there is a physical limit in the interception of raindrops by a given canopy cover. In addition, this effect is complicated by leafdrip which often form large drops that may induce high rates of splash detachment and transport.These results suggest that interaction terms should be included in models for erosion prediction. They not only bring about improvements in the precision of erosion predictions, but also reveal important aspects of the mechanics of erosion that should be pursued. The use of a multivariate orthogonal design has minimised the requisite number of experiments and maximised the amount of information that can be extracted from each experiment.
出处
《地理研究》
CSSCI
CSCD
北大核心
1989年第4期28-36,共9页
Geographical Research
关键词
降雨
击溅侵蚀
多元正交试验
Rainsplash erosion, Multivariate orthogonal
