Mapping grain yield with a combine instrumented for location detection

Abstract

Recently the need to treat agricultural fields as variable entities has been recognized. A practical example is the optimization of fertilizer application rates according to natural variability of soil type. It has also been recognized that variability in grain yield would be an important input in making decisions for site specific treatment. In this research, an automatic combine data acquisition system was developed to monitor in-field grain yield variability. Instrumentation of a grain combine with a grain flowmeter and a microwave location system showed that such a system is technically feasible. The accuracy of the location system was tested. The effect of geometry of the layout of remote transponders on the accuracy of location detection was analyzed. Field tests showed that the location system was capable of tracing the path of the combine with sufficient accuracy. Development of a grain flow model and an analysis procedure allowed construction of detailed grain yield maps. The grain flow model was based on transportation delay and first-order lag. It was found from limited field tests that the transportation delay (13 to 18 seconds) was negatively correlated to the grain flowrate while the time constant of the first-order lag (2 seconds) was independent of it. Sample grain yield maps were created for two small test plots. The variability of grain yield shown in the maps generally agreed with field observations. However, direct comparison with manually taken samples resulted in a mixed success. Such maps can be used to evaluate the variability in crop yield and to determine management decisions accordingly.