Skip to main content
ARS Home » Research » Publications at this Location » Publication #151432
Title: THE VALUE OF ADDITIONAL DATA TO LOCATE POTENTIAL MANAGEMENT ZONES IN COMMERCIAL CORN FIELDS UNDER CENTER PIVOT IRRIGATION

Author
item Heermann, Dale
item Diker, Kenan
item Buchleiter, Gerald
item Brodahl, Mary

Submitted to: European Conference on Precision Agriculture Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 4/1/2003
Publication Date: 6/10/2003
Citation: Heermann, D.F., Diker, K., Buchleiter, G.W., Brodahl, M.K. 2003. The value of additional data to locate potential management zones in commercial corn fields under center pivot irrigation. European Conference on Precision Agriculture Proceedings.

Interpretive Summary: Agricultural producers are collecting yield maps with the new precision agricultural technology but are asking how many years do they need and how can they simply analyze the data. A simplified technique using frequency analysis was developed that is easily understood by the producers. By assigning each data point to a high or low state based on average yield was simple and stable response zones were determined with 5 years data. A three state frequency analysis where yield was assigned to high, medium and low areas allowed a reasonable estimate of the response zones with only one year of data. The response zones are an important step in the delination of management zones for improved management of fertility, weed and water management for the economic benefits of precision agriuculture.

Technical Abstract: The identification of management zones and writing management prescriptions is a major question facing producers to implement precision farming. How much and what kind of data are required is the question of most producers. The exploration of frequency analysis of yield data that can be easily understood by the producer is the focus of this paper. The frequency of high and low yield was combined to estimate the yield response zones. Results showed that the two-state frequency analysis required at least two years of data. The use of a three-state model for analysis of one years data provided an estimate of 25% and 15% and the two-state model with five years of data estimated 39% and 15% to be in the high and low response zones, respectively.