Comparison of Spatial Interpolation Methods of Soil Properties from Pasture Fields.

Advancement in precision agriculture concentrates on arable crops with limited study on grassland due to low economic value and high field variability. An important part of precision agriculture is site specific map. Using inappropriate interpolation methods can result in misleading spatial information leading to potentially wrong decisions. The objective of study was to construct soil map of selected soil properties using different interpolation method and evaluate the relationship between the accuracy of prediction between interpolation techniques. The study was conducted over an area of approximately 198 hectares at Walter C. Todd Agricultural Research Center, Nacogdoches, Texas. One hundred and three soil samples were collected and analyzed for using a grid sampling method with a grid area of 2 hectares. Soil maps were constructed using three interpolation methods; inverse distance weighted (IDW), simple kriging, and completely regularize spline. The accuracy of maps was assessed using cross-validation while the relationships between measured and predicted soil properties were evaluated using linear regression. Spatial interpolation methods used in this study are not superior to each other in estimating soil properties with regard root mean square error. Moderate correlation provides evidence that spatial interpolation is capable of predicting soil properties with acceptable satisfaction, especially in areas with non-uniform terrain.