Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #310748

Title: Monitoring nitrogen status of potatoes using small unmanned aircraft system

item Hunt Jr, Earle
item HORNECK, D. - Oregon State University
item SPINELLI, C. - Collaborator
item TURNER, R. - Collaborator
item BRUCE, A. - Collaborator
item GADLER, D. - Collaborator
item BRUNGARDT, J. - Collaborator
item HAMM, P. - Oregon State University

Submitted to: Precision Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/21/2017
Publication Date: 5/29/2017
Citation: Hunt Jr, E.R., Horneck, D., Spinelli, C., Turner, R., Bruce, A., Gadler, D., Brungardt, J., Hamm, P. 2017. Monitoring nitrogen status of potatoes using small unmanned aircraft system. Remote Sensing. doi:10.1007/s11119-017-9518-5.

Interpretive Summary: Small unmanned aircraft may be an important platform for remote sensing in agriculture; one of the potential applications of small unmanned aircraft would be to fly over a field to determine areas that need additional nitrogen fertilizer to achieve optimum yields. By applying additional fertilizer to those areas, less fertilizer is used reducing costs, and less fertilizer is lost from the field enhancing water quality. Sponsored by a commercial research and development agreement with Boeing Research and Technology, we conducted a fertilization experiment with irrigated potatoes at Oregon State University’s Hermiston Agricultural Research and Extension Center. A small unmanned aircraft was flown frequently over the experiment acquiring image data with bands at near-infrared, red and green wavelengths of light (false-color infrared). During the time when additional fertilizer would be applied, there were differences in leaf area among treatments. The normalized difference vegetation index (NDVI) is sensitive to differences in leaf area to a limit, above which increases in leaf area do not increase NDVI. The leaf area of all plots was higher than the limit so NDVI could not be used to detect potato requirements for additional fertilizer. Plants deficient in nitrogen fertilizer are yellow-green because of a loss of the photosynthetic pigment, chlorophyll, which is detectable by remote sensing with another vegetation index called the Green-NDVI. There were no differences in chlorophyll content among the fertilization treatments, and there were no differences in Green-NDVI. Later in the growing season during potato tuber maturation, both NDVI and Green-NDVI showed large differences among treatments, which was highly correlated with leaf area and chlorophyll content. However, fertilizer application at this late stage does not increase tuber yield or quality. New methods of data analysis are required in order to use small unmanned aircraft for precision agriculture of potatoes.

Technical Abstract: Small Unmanned Aircraft Systems (sUAS) are potential remote-sensing platforms to manage fertilization for precision agriculture. An experiment was established in an irrigated potato field with different N fertilization rates, and a small parafoil was used to acquire color-infrared images over the 2013 growing season. Two spectral indices were determined, NDVI from the red and NIR bands and Green NDVI from the green and NIR bands. Unexpectedly, there were decreases in NDVI and GNDVI of calibration targets with increased camera exposure time, perhaps caused by changes in light spectral quality correlated with low-light conditions. After calibration, both NDVI and Green NDVI were about equal to indices calculated using reflectances from high-altitude aerial photography and the WorldView-2 satellite. During tuber initiation and early tuber bulking, differences in leaf area index and chlorophyll content were only detectable in plots with the lowest N fertilization rates. As expected, N fertilization rates strongly affected both leaf area index and chlorophyll content during tuber maturation. The relationships between NDVI and Green NDVI were hypothesized to differ among N treatments; there were no differences among treatments on either of the two sampling dates, contrary to the hypothesis. However, there was a large difference the NDVI-Green NDVI relationship between the two dates, showing a decrease in N status over time. Average spectral indices at the plot scale may have reduced the ability to detect differences in N status early in the growing season; smaller pixel sizes available from sUAS remote sensing may offer new possibilities for determining N status based on plant cover.