A method to generate and use as-applied surfaces to evaluate variable-rate fertilizer applications

摘要

Variable-rate technology (VRT) has been used by farmers in an attempt to better match inputs to local growing conditions. In theory, VRT minimizes over- and under-application of inputs. However, the limitations and errors of this technology have not been well documented. Further, standard methods for quantifying the application accuracy using VRT do not currently exist, limiting practitioners' knowledge on performance. Therefore, a spatial data model was developed to generate "as-applied" surfaces as a means to evaluate VRT performance of two applicators. The spatial data model uses geographic information system functionality to merge applicator descriptive patterns with a field application file to generate an 'as-applied' surface map representing not only the actual deposition of granular fertilizer but more importantly spatial distribution. Field data were collected and used to validate the spatial model. Comparisons between the actual and predicted application rates indicated moderate to good correlations (0.62 < R < 0.88) for two applicators. Longitudinal offset such as for a global positioning system receiver impacted model performance for one applicator but not the other. A comparison of the actual application rates to the prescription maps illustrated the inconsistency of VRT performance to deliver target rates. Both applicators were only within 10 % of the target rates a small percentage of time (< 45 %) during field operation. Generated as-applied surface maps highlighted errors associated with VRT along with limitations of the technology within site-specific management (SSM). Thus, as-applied surface maps provide a means to properly evaluate VRT while enhancing researchers' and practitioners' abilities to compare and customize SSM approaches.