Skip to main content
ARS Home » Research » Publications at this Location » Publication #168145
Title: ESTIMATING CORN NITROGEN STATUS USING GROUND-BASED AND SATELLITE MULTISPECTRAL DATA

Author
item Bausch, Walter
item Diker, Kenan
item KHOSLA, RAJIV - COLORADO STATE UNIV.
item PARIS, JACK - DIGITAL GLOBE

Submitted to: Proceedings of SPIE
Publication Type: Proceedings
Publication Acceptance Date: 7/6/2004
Publication Date: 11/12/2004
Citation: Bausch, W.C., Diker, K., Khosla, R., Paris, J.F. 2004. Estimating corn nitrogen status using ground-based and satellite multispectral data. Proceedings of SPIE.

Interpretive Summary: Many satellite-based sensors have spatial resolutions too coarse for within field analysis and inadequate repeat coverage for intensive agricultural management. A cooperative study by the Water Management Research Unit and DigitalGlobe, Inc. compared high-resolution multispectral satellite data to ground-based multispectral data to determine if the QuickBird satellite could provide information on a crop's N status equivalent to the mobile ground-based system. Satellite images and ground-based data were acquired on a commercial center-pivot irrigated corn field in eastern Colorado during the 2003 growing season. Results obtained from three clear days of five attempts to acquire satellite images suggest that the QuickBird satellite can be used for within field estimation of corn canopy N status; however, cloud cover over the area of interest is a hindrance to satellite data acquisition.

Technical Abstract: In-season N management of irrigated corn requires frequent acquisition of plant N estimates to timely assess the onset of crop N deficiency and its spatial variability within a field. This study compared ground-based Exotech and QuickBird satellite multispectral data using the normalized GNDVI to produce N status maps of a study site on three days during the corn vegetative growth period. Scale factors to represent N sufficient and N deficient corn were determined for both systems from relationships between the normalized GNDVI and the NSI. A third classification was required for this study to classify areas that exhibited leaf chlorosis that was not caused by N deficiency, i.e., not N related. N status maps generated from normalized GNDVI values showed similar patterns between the two systems for the three corn growth stages (V10, V12, and V15) investigated. However, the extent of the pattern varied between systems. On 2 July (V10), six of six sample sites were correctly classified using leaf N content to indicate plant N status. On the other two days (7 July and 15 July), four of six sites were correctly classified. Two not N related sites were classified as N deficient when leaf N content was adequate. However, due to the similarity of the maps generated from the two systems, QuickBird multispectral data is a useful tool to estimate corn canopy N status and its variability within the field.