Integrating GIS and GPS into a spatially-variable-rate herbicide application system

A spatially variable rate herbicide application system was developed and a site-specific evaluation of its field performance and accuracy was conducted. The system was capable of automatically changing on-the-go the application rate of active ingredients (AI) to meet the requirements of current sprayer field location.;A 4.2 ha field was sampled on an 18.3 m grid for soil texture and percent organic matter (%OM). The soil texture ranged from sandy loam to clay, while the %OM ranged from 0.98 to 2.73 percent. For the preemergence herbicide selected, a herbicide management table was used to determine the appropriate AI application rate for each area of the field depending on spatial variation of field parameter data (soil texture and %OM). For the sampled field, the AI application rate ranged from 3510 mL/ha to 5260 mL/ha.;A geographical information system (GIS) software was utilized to develop a georeferenced map (management map) of field application rates. A direct nozzle injection field sprayer was equipped with a real-time differentially corrected global positioning system (DGPS). A control program was developed to retrieve the desired application rate from the GIS map utilizing position data (latitude and longitude) supplied by the DGPS system. The retrieved application rate was sent, in a voltage format, to a 21X datalogger which was used to change on-the-go the AI flow rate to correspond with the desired application rate at a specific sprayer ground speed and field position.;Results revealed that the DGPS system maintained, on the average, an accuracy of one meter. However, a distance error of location determination produced by the DGPS system reached 30.84 m with a correction message age of 98 seconds.;For the four application rates used in the study, the highest average application rate error (average difference between desired and calculated application rates) and CV values were 2.0 percent and 0.07 percent, respectively for the analyzed samples. The maximum application rate error was 14 percent for 96 percent of the field data points (96 percent of the time). These results showed that the control system was accurate in producing the desired application rate. On the average, the greatest reaction time of the system was 2.2 seconds.;The spatial analysis showed that most application rate errors occurred near transition zones. These analysis also revealed that the contour lines of the calculated application rate maps followed the same pattern and coincide with the management map contour lines. The developed spatially variable rate herbicide application system was found to accurately reproduce the application rate management map in a repeatable fashion.