中文摘要：

选择不同质地紫色土，分别根据土壤颗粒的数量和重量分布计算土壤粒径分形维数，并与Tyler和Wheatcraft土壤水分特征曲线模型的拟合分形维数进行比较分析。结果表明，三种途径获得的分形维数与土壤质地密切相关，均表现出土壤黏粒含量越高，质地越细，分形维数越高。土壤粒径数量分布分形维数（2．98—3．26）和土壤粒径重量分布分形维数（2．73～2．81）存在较大的差异，但二者间却存在显著的线性关系。土壤粒径分形维数与土壤水分特征曲线模型拟合分形维数（2．72～2．84）均存在显著的线性关系，尤其是土壤粒径重量分布分形维数与土壤水分特征曲线模型拟合分形维数数值十分接近。通过建立的土壤水分特征曲线模型分形维数与土壤粒径分布分形维数关系式，结合Tyler和Wheatcraft模型进行土壤水分特征曲线预测，预测值与实测值具有良好的一致性，因而根据土壤粒径分形预测紫色土水分特征曲线是可行的。

英文摘要：

Purple soils different in texture were selected in a study to calculate fractal dimension of soil particle size distribution （PSD） on the basis of accumulative quantity or weight distribution of soil particles. And the fractal dimension of PSD was compared with the fitted fractal dimension of the soil water retention curve （SWRC） model which was developed by Tyler and Wheatcrafl. Results indicate that fraetal dimensions obtained by three different means were all closely related to soil texture, with the higher content of clay, the finer soil texture and the higher fractal dimension. Although the difference between the fractal dimension of accumulative quantity distribution of soil particles （which ranged from 2.98 to 3.26） and the fraetal dimension of accumulative weight distribution of soil particles （ which ranged from 2.73 to 2.81 ） was obvious, a significant linear relationship was observed between them. A significant linear relationship was always observed between the fractal dimension of PSD and the fractal dimension of SWRC （ which ranged from 2.72 to 2.84） , especially between the fractal dimension of accumulative quantity distribution of soil particles and the fractal dimension of SWRC. They were quite approximate in value. Based on the linear function between the fraetal dimension of PSD and the fractal dimension of SWRC, SWRC was predicted through Tyler and Wheatcrafl＇ s model. It was quite obvious that the predicted SWRC agreed well with the measured data. So, estimating soil water retention curve of the purple soil based on the fraetal dimension of PSD is feasible.