194-8 Assessment of the N-Status of Wheat with Remote Sensing.



Tuesday, October 18, 2011: 10:05 AM
Henry Gonzalez Convention Center, Room 007C, River Level

Urs Schulthess1, Dan Long2, Eileen M. Perry3, Doug Weist4 and Klaus Schelling1, (1)RapidEye, Brandenburg an der Havel, Germany
(2)USDA-ARS, Pendleton, OR
(3)Dept. of Primary Industries, Horsham, VIC 3401, Australia
(4)FarmTech, Choteau, MT

RapidEye's constellation of five satellites provides new opportunities for N management of crops. Each satellite's sensor has a band in the red-edge spectral region, which is more sensitive to the chlorophyll content of a canopy than the visible or near infrared red (NIR) bands. While N often is a growth limiting factor, there are also other factors such as water deficiency that limit crop growth. Hence, there are conditions when zones with a low ground cover have a high chlorophyll concentration resulting in the same signal in the red-edge band as a zone with higher ground cover but lower chlorophyll concentration. In order to generate an accurate N recommendation, total N uptake (in the shoots) as well N concentration (in the leaves) ought to be considered. The simple red-edge based normalized difference vegetation index (NDVIre)

[NDVIre = (NIR - RedEdge) / (NIR + RedEdge)]

has shown to be a robust predictor of total chlorophyll per unit surface area, and thus N content in the canopy. To predict tissue N concentration, which often is approximated in the field with SPAD readings, other indices that correct for the total leaf area have shown to be more accurate. Particularly useful is a new planar domain distance index: In step one, a regression between ground cover and NDVIre is fitted. Next, the perpendicular distance of a pixel to the regression line is calculated. Pixels above the regression line have a positive value and an above average N-concentration, while pixels below the line have a below average N-concentration. The further the distance of a pixel to the regression line, the higher or lower is its N concentration. This index is also very useful to validate the efficacy of past fertilizer applications and to accordingly adjust N management zones in the future.

See more from this Division: ASA Section: Climatology & Modeling
See more from this Session: Integration of Remote Sensing, Crop Modeling and ET