采用土壤孔隙表面分形维数预测土壤水分特征曲线  

Prediction of soil water retention curve by surface fractal dimensions

作　　者：刘亚磊[1] 梁杏[1,2] 朱常坤[1] 李静[1] 

机构地区：[1]中国地质大学(武汉)环境学院,湖北武汉430074 [2]中国地质大学(武汉)生物地质与环境地质国家重点实验室,湖北武汉430074

出　　处：《水文地质工程地质》2014年第3期125-130,共6页Hydrogeology & Engineering Geology

基　　金：国家重点基础研究发展计划"973"项目(2010CB428802);国家自然科学基金资助项目(41272258)

摘　　要：采用土壤颗粒体积累积曲线按照不同的粒径分级方法计算表面分形维数,避免了原始计算过程中颗粒密度不变的假设,改进了计算中的缺陷。结合土壤水分特征曲线分形模型—de Gennes模型,预测试样的SWRC。结果发现:不同颗粒分级只会影响表面分形维数的大小;相同颗粒分级情况下,计算得出的分形维数随着土壤粘粒含量降低而减小;采用D2(最大半径为1mm)的分级情况计算出的表面分形维数,预测得出的结果与实测值相差偏大,RMSE≥2.11E-02cm3/cm3只适合粗略估计田间土壤的水分特征曲线;通过D1(最大半径为0.1mm)计算出的表面分形维数,结合de Gennes模型预测水分特征曲线结果与实测值非常接近,RMSE≤0.0105cm3/cm3。研究表明采用土壤颗粒体积累积曲线计算表面分形维数,并预测土壤水分特征曲线是合理的,具有较高的预测精度。Surface fractal dimensions are calculated by soil particle volume accumulation curve through different particle-size grading methods and are used to predict the soil water retention curve（ SWRC） of samples together with the SWRC fractal model. This method avoids the assumption of stable particle density, and improves the former algorithmic method. The results show that different soil particle-size grading methods can impact the dimensions and the dimensions decreases with the decreasing clay content if the same grading method is used. The predictions of SWRC by D2 method is only suitable for a rough estimation in the field as its greater error. While the predictions by D1 method,together with de Gennes Model,lead to perfect fittings with the observed points. Thus,it is reliable to predict SWRC by using the surface fractal dimensions calculated with the soil particle volume accumulation curve.

关 键 词：颗粒体积累积曲线 表面分形维数 土壤水分特征曲线(SWRC) 

分 类 号：P642.68[天文地球—工程地质学]