Using Yield Monitor Data to Determine Spatial Crop Production Potential

Consistent spatial-temporal yield patterns should help determine spatial production potential. Our objective was to evaluate methods for using yield monitor data to develop spatial yield goal maps. Three to seven years of yield monitor data were analyzed for five sprinkler-irrigated cornfields in central and western Kansas. Yield data were block-averaged to 55 m square cells, normalized based on the mean yield, and then used to develop spatial yield goals for subsequent years using six different methods. One method used a uniform yield goal, two methods combined normalized yield monitor data with a uniform yield goal (transitional), and three methods used only normalized yield monitor data from previous years. Methods were evaluated based on their ability to predict the spatial yield pattern of the subsequent year better than the uniform method. Yield monitor data were also segregated based on the temporal CV of each field during the time of the study, and the six methods were evaluated only on the data that were deemed temporally stable. The result of incorporating yield monitor data into yield goals was inconsistent across sites and years. For one site, the two transitional and three yield monitor methods were significantly better predictors of normalized yield. On another field, the uniform method was a better predictor of normalized yield than the yield monitor methods in three of six years, while the yield monitor methods were better than the uniform method in another year. On a third field, the yield monitor method predicted normalized yield better than the uniform method in one of four years with no difference in the other three years. In general, when the correlation coefficient between two years of yield monitor data exceeded 0.70, the methods that incorporated yield monitor data into the yield goal were better predictors of normalized yield than the uniform method. Evaluating these methods using only data from cells where the temporal CV was less than the average temporal CV for the field did not improve the results sufficiently to warrant widespread use of this practice.