Precision Agriculture Research

Autonomous GPS and three differentially corrected GPS (DGPS) sources were tested under static conditions to assess suitability for precision agriculture applications in the vicinity of Hays, KS. DGPS sources included the U.S. Coast Guard Beacon, Wide Area Augmentation System (WAAS), and OmniSTAR, a commercial service. Both WAAS and OmniSTAR broadcast correction messages from geostationary satellites and cover wide areas. The ground-based Beacon currently is more restricted in area of coverage. The standard for absolute accuracy in this study was RTK (Real Time Kinematic) GPS. Horizontal absolute accuracy (referenced to RTK) as measured by the root mean square error was 0.46, 0.49, 1.13, and 1.71 m for Beacon, OmniSTAR, WAAS, and Autonomous GPS, respectively. Corresponding values for horizontal relative accuracy (each mode referenced to its own mean) were 0.41, 0.47, 0.32, and 0.92 m, respectively. The rather significant improvement in relative performance of WAAS compared to its absolute accuracy resulted from the DGPS correction being based on a different geodetic datum than RTK, Beacon and OmniSTAR. This resulted in a northwest quadrant bias for WAAS compared to RTK. Performance of all three DGPS sources met or exceeded design specifications. Availability for the period of the test ranged from 99.2% (Beacon, WAAS) to greater than 99.9% (OmniSTAR), indicating all the DGPS sources tested provided dependable service. ___________________________________________________ Farmers are expanding their use of precision agriculture or site-specific management technologies to optimize inputs and reduce negative environmental impact. Position accuracy required to consistently implement site-specific management normally exceeds that attainable from the autonomous (uncorrected) Global Positioning System (GPS) even with selective availability disabled. Differentially corrected GPS (DGPS) is a common method for correcting some of the inherent errors and provides submeter accuracy. Consistent realization of centimeter accuracy requires survey-grade or Real Time Kinematic (RTK) GPS. Users of DGPS today have access to alternative differential correction sources that vary in cost as well as suitability or reliability depending on location. Numerous comparisons of accuracy of GPS receivers for precision agriculture applications have been made. Recent examples include evaluation of cross-track error in parallel tracking operations (Han et al., 2002) and analysis of errors in GPS-based automatic guidance systems (Ehsani et al., 2002, Gan-Mor et al., 2002). Stombaugh et al. (2002), in search of standards for dynamic GPS accuracy tests, noted the frustrations of assessing relative GPS receiver performance among commercial options using manufacturer specifications because reporting formats vary and because of differences in data filtering or smoothing. They also lamented that accuracy specifications are generally for static conditions, which may have minimal relevance to the mobile world of precision agriculture. They concluded that the Institute of Navigation had developed a static test standard suitable for characterizing performance of GPS receivers used in precision agriculture, but that dynamic test standards were not yet well defined. Quantification of and standards for GPS accuracy are application dependent. Some applications require high absolute accuracy; others depend on high relative accuracy. Buick (2002) provided an analysis of the meaning of those terms as they apply to static and dynamic conditions and to mapping, swathing, and automated guidance applications. Buick also explored the causes of reduced GPS performance external to the GPS receiver.