基于黄土高原区不同质地土壤的室内土壤水分特征曲线系列试验，讨论了Van Genuchten经验模型对土壤水分特征曲线的适用性，重点分析了基于土壤质地的Van Genuchten模型经验参数的非线性预测。研究结果表明：Van Genu⁃chten经验模型对不同质地的土壤均具有很好的适用性，拟合模型的复相关系数R2都大于0．99，说明Van Genuchten经验模型对不同质地的土壤都具有较好的拟合效果；土壤中黏粒含量越大，其持水能力越强，土壤水分特征曲线越陡直，尤其在吸力值小于1000 cm的低吸力段，这种现象更为明显，而土壤中沙粒含量则与黏粒含量表现出相反的性质；利用RETC软件，对Van Genuchten模型经验参数n、α进行基于土壤质地的非线性预测，建立非线性预测模型，模型复相关系数R2均大于0．96，且参数预测值的相对误差在10％左右，所建立的非线性预测模型可信度较高，预测较为准确。%Based on the series of indoor soil water characteristic curve of different soil texture in Loess Plateau region, the research discussed whether the Van Genuchten model was suitable for soil water characteristic curve,and especially analyzed the nonlinear prediction of Van Genuchten model parameters which based on soil texture. The results show that Van Genuchten model is suitable to fit different soil texture of Loess Plateau region, and the square of correlation coefficient(R2) is bigger than 0.97,so it has a very good fitting effect to different soil texture;the more the content of clay in soil, the stronger the ability of holding soil water, and the soil water characteristic curve is more steep, especially at low suction range which the suction is less than 1 000 cm, the experimental phenomena is more obvious, however, sand content in soil shows a opposite nature compared with clay; by using RETC software, the research conducts nonlinear prediction for two empirical parameters nandαof Van Genuchten model based on soil texture and establishes the nonlinear prediction model, at the same time, the complex correlation coefficient( R2 ) of model is bigger than 0.96, moreover, the relative error of parameters between measured value and predicted value are about 10 percent, so the nonlinear prediction model is reliable and has a accurate prediction function.
展开▼
