Skip to main content
ARS Home » Research » Publications at this Location » Publication #133282
Title: ENVIRONMENTAL AND PHYSIOLOGICAL COMPONENTS OF THE COTTON LEAF REFLECTANCE SPECTRUM

Author
item TARPLEY, L. - TEXAS A&M AG. RES. & EXT.
item Sassenrath, Gretchen

Submitted to: Digital Imaging and Spectral Techniques: Application to Precision Agriculture
Publication Type: Monograph
Publication Acceptance Date: 5/1/2003
Publication Date: 10/1/2003
Citation: Tarpley, L., Sassenrath Cole, G.F. 2003. Environmental and physiological components of the cotton leaf reflectance spectrum. Digital Imaging and Spectral Techniques: Application to Precision Agriculture. Chapter 10; pg. 95-109.

Interpretive Summary: Remote sensing offers potential benefits to crop producers through large-scale, frequent, and accurate determination of crop performance over a wide area. Studies to date have focused on developing spectral libraries by correlating known plant physiological functions with modifications to specific regions of the reflectance spectrum. In this methodology, much of the information of the spectral image is lost. The work described here developed a method in which the entire leaf spectral image is utilized in the analysis. By dividing the reflectance spectrum into discrete regions, correlations were developed between physiological function and the entire spectrum. In this way, all of the information gathered from a remote spectrum will be used in development of spectral libraries. It is anticipated that the accuracy and utility of the remote imagery will be greatly enhanced through utilization of the entire spectal image.

Technical Abstract: Principal component analysis of the cotton (Gossypium sp. L.) leaf reflectance spectrum was evaluated for separating and characterizing the environmental and physiological contributions associated with empirical spectral responses. Indices based on reflectances at particular wavelengths would need to be used in combination to simultaneously characterize various contributions to the spectrum, and to do so in a way that allows for change in multiple spectral bands due to an individual contribution. The reflectance spectra of cotton leaves that varied in genotype, age, light exposure, and chemical treatment were pooled for analysis. The first four principal axes were interpretable: two exhibited environmental (light source quality, time of day) variation, and two exhibited physiological (pigment and leaf surface features) variation. The results from corroborative experiments indicate that the natural bases for the principal components were identified. Principal component analysis allows the possibility of i) analytically filtering our environmental variation, ii) partial establishment of the leaf physiological states associated with particular spectral responses, and iii) simultaneous evaluation of leaf environment and physiological state.