Submitted to: Agricultural Research International Conference Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 8/1/2014
Publication Date: 9/16/2014
Citation: Sudduth, K.A., Drummond, S.T., Kitchen, N.R. 2014. Crop sensors for automation of in-season nitrogen application. In: In: Proc. Automation Technology for Off-Road Equipment 2014, pp. 219-226. September 16-18, 2014, Beijing, China. 2014.
Interpretive Summary: Concerns for the environment have made farmers and the public increasingly interested in technologies that help use fertilizers, including nitrogen (N), more efficiently. One way to increase N fertilizer efficiency is site-specific management, where varying amounts of N are applied based on differing crop needs in different areas of fields. Various approaches have been used to determine how much N is needed in these sub-field areas. One approach uses crop reflectance sensors mounted on the front of fertilizer applicators to assess the health of the crop through its reflectance characteristics and control how much N fertilizer is applied. Several commercial crop reflectance sensors are now available, each with a specific set of characteristics. The purpose of this research was to compare commercial sensors used for N fertilizer application control in corn. We collected data with the sensors multiple times over a corn crop and documented differences in their operating performance. For best results, users need to take these differences into account when using sensors to manage N in corn production. The importance of this research is that it makes scientists, farmers, and other users aware of differences between commercially available crop sensors, so that they can use them appropriately in N fertilizer application systems.
Technical Abstract: Crop canopy reflectance sensing can be used to assess in-season crop nitrogen (N) health for automatic control of N fertilization. Typically, sensor data are processed to an established index, such as the Normalized Difference Vegetative Index (NDVI) and differences in that index from a well-fertilized area are used to prescribe a real-time N application rate. Because data comparing the performance of the different sensor designs is lacking, the objective of this research was to evaluate and compare commercial canopy reflectance sensors for N fertilization in corn (Zea mays L.) Two units of each of four commercial sensors – GreenSeeker 505, Crop Circle 210, Crop Circle 430 and CropSpec – were mounted to a high-clearance applicator for field data collection. Data were collected multiple times on corn plots where eight different N rates (0 to 235 kg N ha-1) had been previously applied to create crop reflectance differences. Additional data included Minolta SPAD meter chlorophyll readings and manually measured crop height. Relative NDVI data from GreenSeeker and Crop Circle sensors were highly correlated, while data from those sensors were less strongly related to CropSpec data. CropSpec NDVI tended to be more strongly related to SPAD chlorophyll readings, while the other sensors were more affected by crop height variations. Little variation was seen with any sensor across multiple data collection runs in a single day. The effect of soil darkness in the scene viewed by the sensor was noticeable in nadir sensor data, and became less prevalent where the crop was larger and occupied a larger portion of the scene. The results of this research document the operational differences among these commercial sensors, providing important information to system developers and users of sensor-controlled N application equipment systems.