Making SENSE

Computers have been taking over more chores and decisions for farmers ‌for decades, and one of their next assignments could be determining ‌how much nitrogen to apply to corn, and when to apply it.

A team of University of Nebraska-Lincoln (UNL) researchers has spent the last three years testing whether or not a computer can better answer those questions than farmers and, looking back at their research, the team is excited.

The Project SENSE (Sensors for Efficient Nitrogen Use and Stewardship of the Environment) team, led by UNL specialists Richard Ferguson and Joe Luck, has been working in nearly two dozen central Nebraska corn fields for three summers to determine the amount of nitrogen to apply. They’re using high clearance applicator-mounted crop canopy sensors rather than trusting a farmer’s eye and intuition.

The results from the study showed significant drops in the amount of nitrogen applied and some improvements in financial return.

“We were able to use the sensor system to react to what’s going on in the growing season, using the corn itself as a kind of indicator of nitrogen status,” Luck says.

The project resulted in less nitrogen applied to 80% of fields tested in the three years and increased profits on 53% of those fields.

The sensors detected light reflecting off the corn, using the biomass and the imagery to calculate whether or not a shot of nitrogen was necessary. In short, it’s looking for a yellowing leaf of a plant, and it can do it better than a human.

“The sensors can definitely pick up that stress quicker than we can, no question,” Luck says.

The most dramatic results came in the first two years of the study. In 2015 and 2016 the crop sensor-dictated application used an average of 19.3% less nitrogen compared to what was applied by more traditional methods on strips in the same field. The yields suffered small losses, 2.25% lower in 2015 and 1.51% lower in 2016, but the overall return on the field--factoring in the money not spent on nitrogen--was slightly higher, averaging 1.24% over the two years. That’s about $7.50 a year per acre.

The 2017 numbers bounce back the other direction. The sensors applied more nitrogen than they had in the previous seasons, cutting the savings there from close to 20% to about 8%. That led to a $6.06 per-acre advantage for the farmer-applied strips.

Score one for humanity? Maybe, maybe not. Luck said there wasn’t a clear answer as to why the latest numbers were different. He said farmers in the study were generally among the more efficient with their nitrogen application before the tests started, and, inspired by what they saw, became more so as the experiment wore on.

There are plenty more questions to be answered. While the team is stepping back this particular arm of experimentation, it’s expanding with others, including a drone-based sensor that could be even more attainable to the average farmer who’s not ready for the expense of a high-clearance application system.

“There’s a lot of potential there,” Luck says. “If I could help 80% of the people out there, that’d be a pretty good thing.”

[PF_0818]

Copyright 2019 DTN/The Progressive Farmer. All rights reserved.