Satellite and ground sensors for yield prediction in four crops
Yield prediction is an important step in formulating in-season nitrogen application rates in crops.
Algorithms for yield prediction with ground-based active-optical sensors for use in in-season N application have been developed for several crops, but their use requires that the sensors travel across the entire field. Some means of predicting which fields need supplemental N would help the logistics of in-season N application.
In a recent article published in Agronomy Journal, researchers examined the use of satellite imagery as a yield predictive tool, which would be followed by in-season N application using a ground-based active optical sensor. When satellite imagery was available, the relationships between satellite imagery and yield were comparable to ground-based sensor readings and yield; however, the lack of early-season imagery due to cloud cover or haze limited its practical use.
This study indicates that where cloud-free days are abundant early in the season, use of satellite imagery as a logistic tool could be practical. However, in North Dakota, where May and June are the months where imagery would have logistical use, days with cloud cover and rain predominate. The lack of sunny days in the region opens the possibility for future use of unmanned aerial vehicles as logistic tools in N management.