Appendix I

Global Positioning Systems (GPS)

The application of Precision Agriculture (PA) requires geo-referenced data obtained from yield combines or taken when soil sampling. The location of soil sample points or yield in the field is determined by using satellites and a ground based (or aerial) beacon that knows exactly where it is located. This allows all of the information (crop yield, soil moisture, soil fertility and soil physical parameters) for that sample point to be positioned, by longitude and latitude, on a field map. Various mathematical methods (kriging, inverse distance weighing) are used to estimate the values between the sample points. The sample points have measured values for crop yield, pH, P etc. The maps are used to identify areas of the field that have different levels of nutrients, topography, % clay etc. Interpolating between the points and production of maps of nutrient levels, crop yield etc can be done using software usually based on Geographical Information Systems (GIS).

The Global Positioning System uses satellite signals to define positions on the earth, day or night. However, the signals we use (and everyone else) come from 24 US military satellites that circle the globe. All of these satellites used to have an error built into the signal, however this has been removed. The satellites follow different but constant paths around the earth and at any one time, 4 of the 24 will be "in view" of the GPS receiver anywhere in the world, 24 hours a day. The satellites continuously transmit radio waves at a very high frequency, 1.2-1.5 GHz, and each satellite has an identification code that goes along with its signal.

GPS operation is based on "satellite ranging", that is a GPS receiver on a combine, determines its position by measuring its distance from several satellites in space. The user's GPS receiver measures the time required for the satellite signals to travel from the satellite to the receiver. The receiver uses the time delay to calculate its distance from each satellite. The receiver also has the current position of each satellite in an electronic almanac (a record of satellite positions over time). To get an accurate position point, four satellites are needed for the positioning of the combine. Where all the "spheres" intersect is the location of the combine. This whole procedure is repeated several times a second. Signals from four satellites permit the determination of horizontal position (latitude and longitude), vertical position (altitude) and time.

Back to the military: they receive two signals from each satellite, which allows them to calculate exactly where they are within a few centimeters. Our GPS units receive only 1 signal from each satellite, which typically gives accuracies of 5 to 8 m. However, data can be corrected in real-time. For those of us not in the military, our receivers compensate by using "Differential GPS" (DGPS) or real time correction. The Coast Guard beacon uses the GPS signal to calculate its position. But, because it is fixed, it compares its computed position with its known position, and then broadcasts the difference. Portable GPS units in the field receive that signal, and use it to automatically correct their calculated positions. This corrected GPS reading typically gives locations that are within 1 to 2 m.

Portable hand- held GPS units, which typically cost less than $300, do not receive correction signals, and typically generate locate positions to within 10 to 30 m. There are also other errors in the measurement of location. As well, satellite signals can bounce off of tree branches and buildings. Consequently, hand-held devices are highly inaccurate in the forest.

Precision Agriculture may allow use of variable rate technology (VRT) which allows different rates of lime, manure, potassium etc to be applied depending on the field requirements. A map of application rates is required, and typically, this is based upon a map of soil chemical properties (a map of pH to determine lime application rates). However, more sophisticated analysis of soils, yield, and remote sensing data could be used to generate maps of N fertilizer applications, based upon the "anticipated" corn yield, the "greenness" of the crop, and the soil organic matter. Herbicide application could be based upon airborne images that indicate the presence of weeds. The "art" and the "science" of developing the application rate maps for VRT is an area of ongoing research and development.