Wednesday, January 10, 2007

PAYVAC: Precision Agriculture Yield Variability Analyzer for Cotton

David Zilberman1, Jeanne Reeves2, Edward M. Barnes2, and Calvin Perry3. (1) University of California - Berkeley, 207 Giannini Hall, Berkeley, CA 94720-3310, (2) Cotton Incorporated, 6399 Weston Parkway, Cary, NC 27513, (3) University of Georgia - Tifton Campus, Biological & Agricultural Engineering Department, 2329 Rainwater Rd, Tifton, GA 31794

When asking the question “Will precision technologies pay on my farm?”, one of the first things to consider is the level of within-field yield variability present – if the field conditions are uniform there is no benefit from variable rate management. In order to estimate the level of yield variability needed to signal an economic benefit, a simple decision aid was developed. The decision aid requires a minimal amount of input information: high and low yields within a given field; total production costs; lint price; and size of the field being considered. Assuming yield variation in the field is normally distributed (user supplied yield range is assumed to encompass 95% of the yield variability present), 100 yield observations are generated and then the percentage of the field that has a positive net income is calculated. To make an estimate of potential savings from precision farming, the current analysis assumes areas of the field that are not profitable are removed from production. The predicted precision savings from the decision aid were compared to data from five actual cotton yield maps (one from Arizona, and four from southern Georgia) at a fixed lint price of $0.52 per pound across a range of production costs. In most cases, the difference in predicted precision savings based on the decision aid results were a reasonable estimate of those estimated from the yield maps (errors rarely greater than $4.00 per acre). A more comprehensive model is currently being developed that will allow evaluation of different management scenarios and a preliminary version of that decision aid will be available for demonstration in the workshop.

Poster (.pdf format, 148.0 kb)