南京毛竹林小流域SCS-CN方法初损率λ取值研究

Calibration of SCS-CN initial abstraction ratio of a small watershed in Nanjing bamboo forest

我国是一个水土流失问题相对严重的国家,降雨径流是土壤侵蚀发生的动力,较准确的计算径流量是估算水土流失和水土保持效益评价的关键。SCS-CN模型是美国农业部开发的,用来计算降雨形成的地表径流量的经验模型,由于其简单、高效的优点被广泛运用于不同地区不同立地径流量的计算中。初损率是SCS-CN方法进行流域地表径流预报的基础输入参数之一,影响着径流模拟精度。本文研究确定南京毛竹林小流域初损率取值,为SCS-CN模型在该地区的适用性提供参考。根据毛竹林小流域径流场实测降雨径流资料,采用渐近线法确定CN值,得出λ变化区间,并选取10场降雨资料对参数λ进行率定,使用相关系数R,模型效率系数E,合格率对参数λ进行评价,得出λ=0.25,0.3,0.35相对较好,三者相关系数分别为0.58,0.54,0.49,模型效率系数分别为-9.42,-5.86,-3.14,合格率分别为77.80%,88.90%,77.80%。综合分析,得出λ取0.3在本流域相对较适合,又选取该流域10场降雨资料做模拟径流比对。结果显示:相对于模型理论值λ=0.2时的模拟效果,在该流域λ=0.3时模拟值更加接近实测值,数模型效率系数E由-16.77提升到-1.03,精度大大提高,所以率定λ=0.3为该流域最适初损率。

China has the relatively serious problem of soil and water loss. Runoff is the primary force that results in soil erosion. Accurate calculation of runoff is the key step to predict the soil and water loss,and it contributes to evaluating the efficiency of soil and water conservation practices. SCS-CN model is an empirical model developed by the United States Department of Agriculture,used to calculate the surface runoff of rain. It has the advantages of being simple and efficient. At present,it has been widely applied in different areas and site conditions for runoff calculation. Initial abstraction ratio（ λ） is one of the basic input parameters in the SCS-CN method used to forecast surface runoff,and affects the precision of the model. This study was aimed to determine the value of initial abstraction ratio based on the Nanjing Mosobamboo forest watershed and provide reference of the SCS-CN model in this region. According to the measured data about rainfall runoff in the Moso bamboo forest watershed,the value of CN was determined by using the asymptote method. And then the shifty interval of λ was confirmed from the certain value of CN. We chose the data of 10 rainfall events and used the correlation coefficient ｜ R ｜,model efficiency coefficient E,and the qualification rate to evaluate the parameter λ. The result showed that λ = 0. 25,0. 3,0. 35 are relatively better than other values when conducting calibration test of the parameters λ,with their correlation coefficients 0. 58,0. 54 and 0. 49,the model efficiency coefficient-9. 42,-5. 86and-3. 14, and the qualification rates 77. 80%, 88. 90% and 77. 80%, respectively. A comprehensive analysis showed that λ = 0. 3 was relatively better in the small watershed. Next,we selected data of another 10 rainfall events to simulate the quantities of runoff,and the result showed that the parameter λ = 0. 3 was closer to the measured surface runoff than λ = 0. 2 from the SCS model. The model efficiency coefficient E was improved from-16. 77 to-1. 03. It greatly improves the accuracy of the model. Therefore,calibrating λ = 0. 3 is the optimal initial abstraction ratio of small watershed in the bamboo forest.