It is widely accepted that soils have fractal characteristics in certain spatial scales. However, the exact magnitude of this spatial self-similarity is often neglected when doing fractal related researches. In this study, 43 sets of soil particle size distribution data of the Chao soil in North China Alluvial Plain were collected, and their particle size distributions with 93 levels were determined by laser diffraction analysis method (Beckman Coulter LS 13-320).The spatial magnitude orders of soils can be taken as fractal objectives and their affecting factors were analyzed based on the particle size distribution data. For determining the fractal magnitude of the soils, we used a typical scale method:1) calculate the differential fractal dimensions in all the 93 particle size levels (Di, i=1, 2,…, 93) according to the definition fractal; 2) choose a typical scale which is supposed to follow self-similarity (however, the measured Di values may still fluctuate in a certain range); 3) suppose Di in the typical scale is normally distributed, and calculate the corresponding distribution characteristics; 4) use the distribution characteristics to find out the lower and upper limits of Di which follows the same normal distribution in the whole range from the lower to upper limit (in the study, 0.5%and 99.5%of the accumulation curve were taken as the lower and upper limits, respectively);5) fit Di in the entire fractal magnitude, and the slope is supposed to be the fractal dimension of the soil.The results showed that the Chao soils in the study area mostly had good gradations, and they only had rigorous fractal structure at small to medium particle sizes (large clay to fine silt), but didn’t show spatial self-similarity in coarse sand scale. Soils with more clay particles usually had larger fractal dimensions, and the extents of fractal reduced with the increasing of clay fraction. Soils with good graduations usually had wider fractal extents. The fractal dimension values were significantly affected by soil gradation uniformity coefficient, the small particle fractions (<50 μm except for 5-20 μm) and effective particle size (p<0.01). However, fractal dimension was not significantly correlated with bulk density and saturated hydraulic conductivity. The reason could be that soil bulk density and saturated conductivity are greatly impacted by the soil characters in coarse particle scale;however, the Chao soils are not self-similar and the fractal dimension cannot reflect the soil characters in this scale. Therefore, it is recommended that the fractal magnitude orders of soil particles should be firstly analyzed before applying it to solve soil structure and hydraulic related problems.%土壤在一定空间尺度范围内具备分形特征已经为大多数研究所认同，然而这种空间自相似性的存在区间却容易被忽视。该文利用激光衍射法测定了华北冲积平原黄河沿岸43组壤质潮土的土壤粒径分布数据（共93个粒级），并以此为基础分析了该类土壤颗粒分形特征的存在区间及其影响因素。结果表明，该类土壤大多只是在中小粒径级别均呈现严格的分形特征，在较粗砂粒尺度上不存在空间自相似性。分形特征存在区间会随黏粒含量增加呈现变小的趋势，而级配良好的土壤的自相似性空间范围也相对较宽。土壤颗粒的分形维数会受到土壤级配的不均匀系数、部分粒径较小颗粒的含量（<50μm但不包含5～20μm的颗粒）和有效粒径等的明显影响；但该分形维数与土壤容重和饱和导水率的关系却并不明显，这可能与该土壤在与结构和水力学性质关系密切的大颗粒尺度上并不具备分形特征有关。因此，研究土壤颗粒分形与土壤结构和水力学性质的关系时，应特别注意分形特征存在的空间尺度。
展开▼
