Abstract

In order to explore the impacts of scale on soil variability, a study in a 120 × 40 m agricultural field was carried out. Spatial variations of soil water content (SWC) and electrical conductivity (EC) were analyzed with all the data measured under 5 × 5 m sampling grid. Then, resampling technique was employed with changing sampling extents and spacing, and spatial variation of each scenario was also determined. The results showed that SWC and EC presented weak or moderate spatial variability, and showed apparent mosaic pattern at the field scale. For all data sets the number of measurement (n = 207) effectively collected was much higher, making the field mean soil moisture and salt content data set very accurate. All the spatial indices measured increased with various degrees with increasing extent. However, spacing did not affect coefficient of variation and had inconsistent effects on the geo-statistics parameters, such as the nugget, sill and the correlation length for SWC data and EC data.

Key words: Soil water content, electrical conductivity, spatial variability, scale-dependency, sampling design.