Analysis of Site-Specific Adjustment Applied to On-The-Go Soil Sensing Data for Agronomic Use

Successful variable-rate applications of agricultural inputs such as lime rely on the quality of input data. Systematic grid soil sampling is the most common method used for creating variable rate prescription maps. The insufficient number of point measurements usually obtained using this method has been primarily responsible for the typical inaccuracies seen in lime prescription maps. To increase sampling density, on-the-go sensing technology was developed for the mapping of soil pH and other relevant attributes. In this study, five fields in eastern Nebraska were mapped using both on-thego sensing technology and systematic grid sampling. Ten calibration points per field were selected to relate sensor data and laboratory test results (soil pH and buffer pH). Also, at least nine validation points per field were used to compare soil pH, buffer pH, and lime requirement estimates predicted using different mapping strategies and those derived from the laboratory measurements. The data collected were used to compare three soil acidity management scenarios: 1) uniform rate liming; 2) variable rate liming based on systematic grid sampling, and 3) variable rate liming prescribed using sensor-based mapping. In general, sensor-based maps were better predictors of soil pH, buffer pH, and lime requirement than field average or grid-based maps. Because the individual calibration points play an important role in the liming requirement outcomes, an analysis of these sites was conducted comparing three sizes of calibration sets