|Sudduth, Kenneth - Ken|
Submitted to: European Conference on Precision Agriculture Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 9/9/1997
Publication Date: N/A
Technical Abstract: The spatial relationship between crop yield and soil and topographic parameters was modeled using several methods. The goal was not only to estimate yield based on the other spatial data but also to understand the shape of the yield response to changes in measured limiting factors. The approach involved empirical analysis of multivariate spatial datasets which hincluded soil parameters, topography, and multi-year yield data obtained o research fields in Missouri. Projection pursuit regression, a multivariate nonparametric method, was able to produce estimated yield maps which were similar to actual yield maps. The estimated yield response curves generated by this method appeared to successfully model major yield-limiting factors. For the fields studied, yield responses to soil and topographic parameters were highly variable between years due to differences in climatic conditions.