|Tarpley, L - MISSISSIPPI STATE UNIV|
|Reddy, K - MISSISSIPPI STATE UNIV|
Submitted to: Journal of Environmental Quality
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: June 2, 2002
Publication Date: September 1, 2002
Citation: READ, J.J., TARPLEY, L., MCKINION, J.M., REDDY, K.R. NARROW-WAVEBAND REFLECTANCE RATIOS FOR REMOTE ESTIMATION OF NITROGEN STATUS IN COTTON. JOURNAL OF ENVIRONMENTAL QUALITY. 2002. v.31p.1442-1452. Interpretive Summary: Nitrogen (N) often limits yield in upland cotton. Wavebands of reflected light sensitive to N stress should aid in applying remote sensing technologies to site-specific nutrient management. Further, reflectance- based difference in plant N status will be more widely accepted if both leaf-level and canopy-level measurements provide nearly equivalent spectral linformation. To test this hypothesis, we grew cotton outdoors in large pots and used a computerized drip-irrigation system to provide prescribed amounts of water and nutrients to each pot. The spectra of reflected radiance from 350 to 950 nanometers (nm) were acquired for both single leaves and canopies at midday under incident solar radiation. Results suggest spectrally narrow reflectance wavebands involving the red-edge (680 to 720 nm) and violet or blue regions of the spectrum (400 to 450 nm) should enhance our capability to remotely estimate N status in cotton using an imaging or non-imaging sensor. Crop reflectance data can be used either as input to a decision support system or to deploy variable-rate application equipment for more accurate and timely management of water and nutrients. The potential for remote sensing to enhance farm fertility management will best be achieved if the airborne-image data are processed using a well calibrated spectral library, giving results that are readily interpretable by the producer or consultant when making management decisions.
Technical Abstract: Decreased nitrogen(N)availability can dramatically alter leaf pigmentation and plant structure in cotton (Gossypium hirsutum L.), and has thus been shown to alter the spectra of reflected radiance. An N-deficiency study was conducted to determine ratios of leaf/canopy reflectance mesured in narrow bandwidths (1.5 and 5.0 nm) that correspond most strongly with changes in leaf N status in cotton. The five N treatments were half- strength Hoagland's nutrient solution (control), 20% of control N following first flower bud (square), 0% and 20% of control N following first flower, and 0% of control N following fruit (boll filling). Chlorophyll and leaf N concentration decreased significantly across the five N levels. In general N stress led to increased reflectance at about 695 +/- 5 nm and decreased reflectance at about 410 nm. Canopy reflectance ratios that most strongly indicated plant stress were R415/R695, R415/R585, R710/R360, and R755/R700, ,explaining 56 to 72% of the variation in leaf N status. Multiple linear regression identified R437.52 and R610.56, spectral regions within +/- 25 nm of those obtained using correlation, and indicated the best four- variable, canopy waveband model explained 84% of the variability in leaf N concentration. Results suggest remote sensing of N status, and thus stress in cotton is possible using narrow band relfectance ratios that involve the violet or blue region of the spectrum (400 to 450 nm) and the more commonly featured red-edge (680 to 720 nm) region.