【摘要】 三维属性建模是利用有限的采样数据,通过插值或模拟的方法来重构地学属性在三维空间中的分布.将Kriging方法推广到三维空间,从而演化为三维Kriging方法,可以为三维属性建模提供可靠的手段.而三维Kriging方法面临的一大难题就是各向异性变异函数的套合.提出了一种简单通用的三维空间变异函数的套合方法.该方法以空间坐标基的变换为基础,在套合时充分考虑轴向上变异差异的影响,并由此提出各向异性变化率的概念;论证了套合方法的可行性,并通过地下水水质三维属性建模的实例对该方法进行了有效的验证.更多还原

【Abstract】 3D property modeling is hot nowadays in geoscientific researches,the essence of which is mapping geological properties in 3D space from limited samples by interpolation or simulation methods.By extending Kriging to 3D space which evolves to the so-called 3D Kriging,a reliable method can be provided for 3D property modeling.However,one of the most difficult problems in 3D Kriging is the nested overlap of anisotropic variograms.In this paper,a simple but universal method is proposed for the nested overlap.Based on the transformation of coordinate basis,this new method takes into full consideration the difference of variation on each axis direction and puts forward the concept of anisotropy ratio.The feasibility of the new method is demonstrated and testified by a case study about 3D property modeling of groundwater quality.更多还原