Skip to main content
ARS Home » Southeast Area » Mississippi State, Mississippi » Crop Science Research Laboratory » Genetics and Sustainable Agriculture Research » Research » Publications at this Location » Publication #169392

Title: MID-SEASON COTTON YIELD VARIABILITY ESTIMATION WITH REMOTE SENSING

Author
item IQBAL, JAVED - MISSISSIPPI STATE UNIV
item Jenkins, Johnie
item THOMASSON, ALEX - MISSISSIPPI STATE UNIV
item McKinion, James
item Willers, Jeffrey
item WHISLER, FRANK - RETIRED MISS STATE UNIV

Submitted to: Agronomy Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 6/1/2005
Publication Date: 11/1/2005
Citation: Iqbal, J., Jenkins, J.N., Thomasson, A., McKinion, J.M., Willers, J.L., Whisler, F.D. 2005. Mid-season cotton yield variability estimation with remote sensing. Agronomy Abstracts. 2004 CDROM.

Interpretive Summary:

Technical Abstract: Agriculture remote sensing offers a potential source of data for crop management. The objective of this study was to estimate cotton (Gossypium hirsutum L.) lint yield using remote sensed data and compared with data from yield monitor. Multispectral aerial images with four bands (near infrared, red, green, and blue) on several dates were collected with a 0.50-m spatial resolution. Images were used to compute vegetation indices, such as ratio vegetation index or RVI, normalized difference vegetation index or NDVI, and green vegetation index or GVI. However, the best coefficient of correlation (r2 = 0.91) was found between NDVI and crop yield. In 2002 a linear model between NDVI and cotton lint yield was derived and was used in a GIS to convert field based NDVI maps of 2002. The same linear model was used on another field in 2003 growing season to convert NDVI map into cotton lint yield map. A simple raster differencing technique was used to create residual cotton lint yield maps for each season. The frequency distribution of residual yield maps showed a mean and standard deviation of 15 +/- 192 kg ha-1 and 89 +/- 213 kg ha-1during 2002 and 2003 growing seasons, respectively. The suggested methodology could be used for other crops and would enable producers to better monitor their crops during the season.