AIRBORNE MULTISPECTRAL IMAGERY FOR MAPPING VARIABLE GROWING CONDITIONS AND YIELDS OF COTTON, GRAIN SORGHUM, AND CORN

Abstract

Airborne remote sensing is becoming increasingly useful for mapping plant growth and yield variations in precision agriculture, but operational methodologies are not well developed nor tested. The main objective of this study was to integrate airborne multispectral imagery, ground observations, global positioning systems (GPS), geographic information systems (GIS), image processing, and yield monitoring for mapping spatial variations in plant growth and yield. A 30.4–ha dryland field that was divided into two conventional tillage strips and two conservation minimum tillage strips was the study site. The field was planted to cotton (Gossypium hirsutum L.) in 1996, to grain sorghum (Sorghum vulgare Moench) in 1997, and to corn (Zea mays L.) in 1998. Airborne color–infrared (CIR) digital imagery was acquired from the field on three dates in each growing season and ground observations (plant populations, height, yield) were made at 29 sites within the field. A yield monitor was also used to record yields at harvest for grain sorghum and corn. These images and the ground measurements were integrated within a GIS to document, interpret, and map within–season and across–season plant growth and yield variability. The images clearly revealed plant growth patterns within and across the three growing seasons as well as differences between the two tillage systems. Yields of cotton, grain sorghum, and corn were related to the image data for the three spectral bands and four vegetation indices (two band ratios and two normalized differences) extracted at the 29 sampling sites. Regression equations for yield as a function of a spectral band or a vegetation index for each crop were developed and the best equations with R 2 values of 0.57, 0.59, and 0.76 for cotton, grain sorghum, and corn, respectively, were used to estimate the yields for each of the approximately 30,000 pixels. The yield maps generated from the image data based on the regression equations corresponded closely with yield monitor data maps. Recommended operational procedures are summarized. This study illustrates practical ways to integrate airborne digital imagery with spatial information technology and ground observations to map plant growth conditions and yield variations within crop fields.