Interpolation type and data computation of crop yield maps is important for precision crop production

Technological advances in precision agriculture in the last two decades have made yield monitoring and mapping an economically feasible option or practice for farmers. Differentially corrected Global Positioning System (GPS)-equipped yield monitoring system on a combine allows collection of georeferenced yield data which when coupled with a geographic information system (GIS) can generate yield maps via several interpolation techniques. Scientists and practitioners have reported to use multiple different types of interpolation techniques to process yield data. However, one of the aspects that still need to be elucidated is the influence of the different interpolation methods on the quality of the resulting thematic yield maps. The objective of this study was to investigate the influence of three interpolation methods (i.e., inverse of distance, inverse of square distance, and ordinary kriging) commonly used in developing yield maps. An index for the comparison of errors (ICE) was proposed to provide an objective criterion for selecting an experimental variogram model to use with the kriging. Results indicate that inverse distance squared performed slightly better in predicting yields than either inverse distance or ordinary kriging. With a mean absolute difference varying from 0.04 to 0.32 t ha^?1 corresponding to a relative deviation from 1.20 to 7.53%, the management decisions can differ in some cases based on the type of interpolation implemented.