Skip to main content
ARS Home » Research » Publications at this Location » Publication #153679


item Read, John
item Willers, Jeffrey
item Jenkins, Johnie

Submitted to: Society of Photo-Optical Instrumentation Engineers
Publication Type: Proceedings
Publication Acceptance Date: 9/4/2003
Publication Date: 10/1/2003
Citation: Read, J.J., Iqbal, J., Thomasson, J.A., Willers, J.L., Jenkins, J.N. 2003. Remote sensing in dryland cotton: relationship to yield potential and soil properties. Proceedings Ecosystems Dynamics, Agricultural Remote Sensing and Modeling, and Site-Specific Agriculture. 5153:61-72.

Interpretive Summary: Studies were conducted in a field of non-irrigated cotton to obtain information about soils and field elevation, and their effects on vegetation reflectance, an estimator of plant vigor. Our objective was to determine the relationship between normalized difference vegetation index (NDVI) and cotton lint yield at 24 field locations, and use the relationship to estimate yield from a high-resolution aerial image of NDVI. Results indicate NDVI acquired at peak flowering stage in late July can be used to produce field maps of plant height and leaf area for that same time frame, as well as estimated lint yield in September. Results support evidence of a strong relationship between cotton yield potential and photosynthetic leaf area during fruit-filling period. Across 24 locations, yield increased as soil elevation and % sand content decreased, supporting evidence that cotton growth is responsive to soil moisture conditions. For instance, strong spatial correspondence was obtained between classified maps of yield and soil bulk density or plant available water content. Results indicate growth and lint yield in cotton were associated with NDVI values acquired at mid-season, during the peak of flowering and fruit development. Results suggest that remote sensing of cotton vigor in conjunction with information on soil-landscape topography can be used in site-specific crop management.

Technical Abstract: Remote sensed data is a potential source of information for site-specific crop management, providing both spatial and temporal information. Our objectives were to determine the spatial variability in cotton growth based on multispectral image data, and in soil physical properties based on kriging techniques in three soil horizons in a 42-ha field. Image data of soybean from 2000 was used to construct a map of normalized difference vegetation index (NDVI) and, hence, to establish 12 locations that comprised four sites in each low, medium and high NDVI class. Measurements were obtained about every other week for leaf area index, site-specific hyperspectral reflectance, and visible/near-infrared (NIR) digital imagery at 1-2 m ground spatial resolution. Field plots in the image data were overlain with a 144 -m2 area in order to extract average radiance values for calculation of plot NDVI. Plant height, leaf area index, and lint yield were closely associated with NDVI maps and with NIR band values acquired from either an aircraft or handheld (GER-1500) sensor during peak bloom in mid July. Results indicate NDVI and NIR reflectance measures obtained in July can be used to produce estimated field maps of plant height, leaf area index and end-of-season yield, and thus offer important mid-season management tools for site specific farming, especially in dryland cotton.