Skip to main content
ARS Home » Research » Publications at this Location » Publication #127971
Title: REMOTE SENSING: A NEW APPROACH TO VARIABLE RATE HERBICIDE APPLICATION

Author
item ALDERKS, DAVID - UNIV OF ILLINOIS
item SPRAGUE, CHRISTY - UNIV OF ILLINOIS
item WAX, LOYD

Submitted to: North Central Weed Science Society US Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: 12/14/2001
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: In 2000, a field scale study was set up in a producer's soybean field near Champaign, Illinois. The objective was to determine if remote sensing could be used to detect different weed populations in the field and develop a variable rate herbicide applications map from the resulting images. Prior to herbicide application, weed densities and heights were collected from 66 4-m2 random points for ground truth and accuracy assessment. Airborne multi-spectral imagery was taken at a 4-m2 resolution and merged with 1-m2 panchromatic imagery to simulate pan- shaped IKONOS satellite data. An unsupervised classification on the imagery was performed to generate a herbicide application map. Ground truth data and class statistics were used to group the weed infestations into three classes, low, medium, and high. Three glyphosate rates of 33%, 67%, and 100% of the labeled use rate were assigned to the low, medium, and high classes, respectively. An accuracy assessment indicated the classification was very successful at delineating the weed infestations. Variable rate applications based on the herbicide application map were compared with a full application rate of glyphosate for weed control. Results from these plots indicated that there were no differences in weed control and yield between the two application methods.