Assessment and visualization of spatial interpolation of soil pH values in farmland

Site-specific farming entails fine-scale detection of field parameters that affect yield coupled with directing appropriate management inputs to select areas that improve field-scale cropping system profitability. Currently, limited technologies are available to evaluate spatial variability in soil properties on a fine scale (submeter resolution). Therefore, information is typically generated by collecting discrete samples and utilizing spatial interpolation to estimate data for the unsampled locations. In this study, soil pH samples were collected from a 12.15 ha agricultural field in northwest Missouri using two grid-sampling regimes: 0.11 ha with 110 samples and 0.98 ha with 12 samples. Three spatial interpolation methods (inverse distance weighted, spline and kriging) were tested to evaluate the effects of interpolation on unsampled locations. In addition to quantitative validation evaluations, results were also assessed by 2D visualization and 3D visualization. Although each assessment approach provided useful information, the inverse distance weighted technique overall better-estimated soil pH values as determined by a combination of all three approaches.