Spatial Variability of Soil Moisture Content and Reasonable Sampling Number in Cluster-Peak Depression Areas of Karst Region

In the cluster-peak depression areas at northwest Guangxi,soil moisture content at surface layers(0~5,5~10,10~20,20~30 cm) was measured in a rectangular area(90m×40m).The spatial variability construction and distributing pattern were discussed by using the traditional statistics and geo-statistics,and the reasonable sampling numbers were calculated under different believable level and precision for all layers.The results showed that soil moisture contents at all layers had good semi-variance structure and its spatial pattern came out as obvious patches.Soil layer of 0~5 cm had medium spatial correlation,and the other layers had strong one.And the extent of spatial variability of soil moisture changed with layers.The first three soil layers could be well expressed by exponential models,but the last layer(20~30 cm) was simulated with spherical model.The soil moisture spatial variability in this region mostly contributed to the influencing factors,such as topography,micro-physiognomy,rainfall and vegetation from the analysis of semi-variance and contour maps.Moreover,the traditional statistics could figure out the sampling numbers of soil moisture distinctly,but in practical applications,the spatial construction of soil moisture should also be considered.