The Spatial Analysis of Yield Data

Abstract

Yield data are now recorded automatically for a wide variety of crops including cereal grains, oilseeds, fiber, forage, biomass, fruits and vegetables. Yield monitors for grain crops were developed in the 1980s and commercialized by 1992 before global positioning systems (GPS) had become fully operational for civilian use. Since the initial agricultural use of GPS, farmers have made intensive use of yield monitors. This Chapter describes how yield data from monitors must be calibrated and how measurement errors can be addressed. Yield data can be used to target crop and soil investigations, nutrient applications and for on-farm experiments. Cautions for the use of yield data for decision making are considered; important amongst these are the difficulties associated with spatially correlated data for traditional statistical analyses and the alignment of different spatial layers. This Chapter uses spatial statistics rather than only geostatistics; the latter comes within this broader category. Some links between the different approaches in spatial analysis are illustrated. Spatial statistics lends itself better to econometrics than does geostatistics, and the importance of the economics of precision agriculture is discussed. The methods described are applied to a case study of soybean data.