Use of Unsupervised Clustering Algorithms for Delineating Within-Field Management Zones

USE OF UNSUPERVISED CLUSTERING ALGORITHMS FOR
DELINEATING WITHIN-FIELD MANAGEMENT ZONES*

C.W. Fraisse, K.A. Sudduth, N.R. Kitchen, and J.J. Fridgen

SUMMARY

Variable rate application of crop inputs has become common practice in certain areas of the country. A potential failing of site-specific management is that the decision rules for varying inputs are based on recommendation algorithms obtained from multiple locations. Delineation of field areas that are subject to a unique combination of potential yield-limiting factors would allow more accurate management of inputs. Unsupervised classification of topographic attributes and soil electrical conductivity was applied to two fields for delineating potential management zones. Evaluation of the zones using yield data indicated that up to a point, within-zone yield variance generally decreases by increasing the number of zones. The ideal number of zones to use when dividing a field may vary from year to year and is mainly a function of weather and the crop planted

* Presented as Paper No. 993043 at the 1999 ASAE International Meeting,
Toronto, Ontario Canada, 18-21 July 1999.

To view the entire document, click the PDF button on the left.

Research Publications