Since the early days of satellite imagery, crop scientists and mathematicians have been studying those pictures with an eye toward getting good data. The images — called the Normalized Difference Vegetative Index, or NDVI — can be used along with complex models to predict crop yields from afar.

Kevin Price, Kansas State University professor of agronomy and geography, was an early adopter of the models. “I used these data to predict yields on eight major row crops in the U.S.,” he says. 

Price, who has been gathering this NDVI, or vegetative health index, data since 1989, says he has more than 1,200 images on hand and can build models for every 250-acre area in the continental U.S.

The key is making this information available to farmers in the future, an effort Price is undertaking with a software development project that would allow agronomists to look at trends for any area in the U.S.

Add in the rise of unmanned aerial systems that provide high-resolution farm images, and there’s an additional benefit of the NDVI work. Price says he applied his vegetative health model to cornfield data from an unmanned aerial system and explained more than 92% of the variation in the field’s yield.