本文以田间试验资料为基础,研究田间土壤含水率的空间变异性与空间结构。对1m土层不同深度的含水率与1m土层平均含水率作了统计、自相关与半方差分析,表明田间土壤存在着空间变异性。1m土层平均含水率的变差系数C_v小于各不同深度含水率的C_v值。资料还表明含水率在平面位置上相关性不显著。将田间含水率作为空间变异的随机变量,可计算其符合要求的置信水平P1和估值精度的合理取样数目。本文提供的取样数目表亦可供确定合理取样数目时参考。

This paper deals with spatial variation and spatial structure of soil一water content based on the test data in the field. 100 sampling sites were arranged along a line with one meter interval locat-ed at the Experimental Station of Guan County in Shandong Province. 505 samples were taken at depths of lOcm，3Ocm，5Ocm，7Ocm，9Ocm，89 samples for bulk density were taken in correspon-deuce with the samples for soil-water content. The analysis of statistical parameter，autocor-relation，and semivariogram have been done for the soil-water contents at different depths the avarage 0-100cm soil-water contente Field data obtained form the study area show that the co- efficient of soil-water content CVs is about 0. 1一0. 25. The values are similar to the CVs from 78 series(5000 samples)，which we were taken form Shanxi，Inner Mongolia，and Hebei Provences. The coefficient of avarage soil-water content CV at 0一1 OOcm is smaller than that of the soil-water content CVs at different depths. 1t is possible to calculate soil-water storage capaci-ty for one meter depth by using average soil-water content and average bulk density. The spatial relationship among the field soil-water contents is of little Significance. Samples which have a certain distance can be treated as random variables. Then the number of samples necessary for various probability level (or confidence level) PI and specified limit (or certain es- timation error )\Can be determined. A method of computer-similitude have been applied to check the necessary number of samples otained form the table. In general ,sampling for soil-water content with specified limit of A=10一l5% at confide-nce level of PI=90一95% is acceptable in practic.