Skip to main content
ARS Home » Research » Publications at this Location » Publication #136744


item Greenberg, Shoil
item Everitt, James
item Sappington, Thomas

Submitted to: Transactions of the ASAE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/3/2003
Publication Date: 8/26/2003
Citation: Yang, C., Greenberg, S.M., Everitt, J.H., Sappington, T.W., Norman, J. 2003. Evaluation of cotton defoliation strategies using airborne multispectral imagery. Transactions of the ASAE. 46(3):869-76.

Interpretive Summary: Traditional methods for evaluating cotton defoliation strategies are based on visual observations and ground measurements. This paper presents a remote sensing-based method for evaluating the effectiveness of different cotton defoliation strategies. Results from two cotton fields indicated that airborne imagery permitted both visual and quantitative differentiation among treatments. This remote sensing-based technique has potential for more effective and efficient evaluations of harvest aids.

Technical Abstract: Visual observations and ground measurements are commonly used to evaluate cotton (Gossypium hirsutum L.) harvest aids for defoliation, boll opening, and re-growth control. This paper presents a remote sensing-based method for evaluating the effectiveness of different defoliation treatments. Field experiments were conducted on two cotton fields in south Texas in 2001, in conjunction with a study on the effects of defoliants alone and in combination with insecticides on boll weevil mortality. Eight treatments (one control and seven combinations of defoliants and insecticides) with three replications were assigned across 24 experimental plots in a randomized complete block design in each of the two fields. Airborne color-infrared digital images were obtained from the first field six days after chemical application and from the second field on four equally-spanned dates beginning on the day of application. Ground reflectance spectra and plant physical data, such as number of leaves, were collected on selected sites within each plot. The reflectance spectra effectively separated different levels of defoliation, but a large number of spectra were required to obtain reliable results. The airborne multispectral images permitted visual differentiation among the treatments as early as three days after the chemical application, though the images collected six days after the application revealed the most significant differences among the treatments. For quantitative analysis, the green, red, and near-infrared bands and the normalized difference vegetation index (NDVI) were used as spectral variables to compare the differences among the treatments. Multiple comparisons showed that spectral variables differed significantly among some of the defoliation treatments. These results indicate that remote sensing can be a useful tool for evaluating the effectiveness of cotton defoliation strategies.