空间变异研究中最为常用的克里格插值法能够较为精确的估值,但由于其低通滤波性,往往会对较高数值的异常点产生低估。而多维分形方法正是强化局部插值结果,刻画局部奇异性的有效工具。应用多维分形地统计学方法于铜陵地区土壤中Cd元素的空间分布研究,结果表明,该地区土壤中Cd元素有较强的空间自相似性,多维分形方法合理地强化了局部区域的估值结果,可以较好的表达土壤元素空间分布的局部奇异性。 The most commonly used Kriging interpolation method in spatial variability research can estimate more accurately, but due to its low-pass filtering, it often underestimates the abnormal values ​​of higher values. The multi-dimensional fractal method is to strengthen the local interpolation results, an effective tool to describe the local singularity. The spatial distribution of Cd in soils in Tongling area was studied by multidimensional fractal geostatistics. The results showed that there was a strong spatial self-similarity of Cd in soils, and the multidimensional fractal method reasonably intensified the estimation results of local regions , Can better express the local singularity of the spatial distribution of soil elements.