【摘要】 将稳健统计学(RS)引入地质统计学(GS),研究土壤水盐特性空间变异。重点研究了稳健型变差函数γR(h)的逼近原理和推求方法,本方法应用在大型灌区土壤水盐时空变异性预测和监测的结果表明,对于实际数据分布不严重偏离理想分布的条件下的中小样本,可以获得分布稳健和数据稳定的γR(h)及其估计量,这对于改善变差函数的结构性和理论模型的主要参数有明显效果。更多还原

【Abstract】 The Robust statistics is combined with Geostatistics to study the spatial variability of water and salinity in soil. The principle for approximation of Robust variogram and its method are studied. The application of this method to forecasting and monitoring of spatial variability of soil water-salinity in large-scale irrigation area shows that the Robust distribution and stable vriogram can be obtained if the skewness of the statistics distribution of the samples with moderate-size or small-size does not seriously deviate from ideal distribution. This is effective for improving the structure of variogram and the parameters of its theoretical model.更多还原