黄河三角洲地区土壤盐分的空间变异性及其CoKriging估值

Spatial Variability of Soil Salinity in the Yellow River Delta and Its Estimation by CoKriging Method

以黄河三角洲地区典型地块为研究区,运用经典统计学和地统计学相结合的方法,研究不同深度土层盐分的空间变异特征,绘制各土层盐分的随机性和结构性的半方差图以及空间分布图。分析表明:受结构性因素和随机性因素的共同作用,各土层盐分均具有中等的变异强度和空间相关性。Kriging插值结果表明:试验区各土层盐分均表现为条带状和斑块状分布,且各土层盐分的空间分布在一定范围内均存在着空间上的相关性。微地形和气候条件是影响表层土壤盐分空间分布的主要因素,地下水性质是主导深层土壤盐分空间分布的主要因素。引入表层土壤盐分,采用CoKriging方法对深层土壤盐分进行估值,可提高估值精度,其估计方差减少百分数达167.36%。

In this paper, the spatial variability of soil salinity at all depths in some typical fields in the Yellow River Delta is lucubrated by using statistics and geo-statistics. The study area （37°33′- 37°34′N, 118°47′- 118°50′E） is located in Yong＇ an Town, Kenli County, Shandong Province, and it belongs to the temperate continental monsoon climatic zone, the spatiotemporal distribution of precipitation varies, precipitation occurs mainly during the period from July to August and its proportion occupies 70% of the annual precipitation, and the ratio between annual evaporation and precipitation is about 3. Spatial distribution figures and semi-variograms, which can be used to explicitly explain the random and structural variability of soil salinity, are charted. The results show that there is a moderate spatial variability and a spatial correlation in soil salinity at all depths, and the spatial distribution of soil salinity is jointly affected by structural and random factors. All the spatial distribution figures of soil salinity interpolated by Kriging interpolation show apparently that the distribution of soil salinity in the study area is belt and patch-shaped, and there is a spatial correlation between the distribution figures of soil salinity at different depths in certain extent. The spatial distribution of salinity in topsoil is mainly affected by micro-topography and climatic conditions, and that in deep soil is mainly affected by groundwater properties. The values of salinity in deep soil can be estimated by CoKriging method using the val- ues of salinity in topsoil, thus the estimation precision of salinity in deep soil can be significantly increased, and the estimated variance can be decreased by 167.36 % compared with that estimated by Kriging method.