Applying nitrogen properlyScientists study height, optic results for corn predictions.
By: Mikkel Pates, Agweek
FARGO, N.D. — Imagine a future in which farmers use scientifically gathered corn crop height data to color intensity ratings and decide whether to add mid-season nitrogen.
North Dakota State University researchers say it’s not that far away. An optical sensor study for predicting corn yields was one of the topics when the NDSU Extension Service on Jan. 15 sponsored “Soil and Soil/Water Training” at the Fargodome in Fargo, N.D. The training offers required continuing education credits for more than 200 crop consultants who attended.
Lakesh Sharma, at NDSU since 2011, presented early results for a corn N-rate study he’s working on for a doctorate study in conjunction with David Franzen, an NDSU Extension soils specialist. Sharma laid out data that shows how Greenseeker and Holland Crop Circle brand technologies may be able to predict corn yields and, therefore, N-rate side-dress application requirements in corn.
Franzen explained that one sensor uses red ratings, divided by near-infrared ratings, which is also known as normalized difference vegetative index (NDVI). Satellites have long been used to estimate things such as a wheat crop in a foreign country.
“It’s kind of a two-dimensional thing, so it’s looking at biomass on a two-dimensional scale,” Franzen says. “It was originally developed for winter wheat, Bermuda grass — things that lie lower to the ground. What we’re doing is using it on corn.”
Height plus optical
The new study uses two different technologies, Franzen says. One is the NDVI. The second is the Holland Crop Circle Sensor — a tool developed in Nebraska. That has the NDVI, but also a lens that gives what is called a “red edge.”
“The ‘red edge’ NDVI is a kind of a proxy for the green that we see,” Franzen says. “It goes beyond the two-dimensional leaf area of a wheat crop and it puts into play something we would see going out into the field but it quantifies it in a way that our brain can’t, really.”
The first part of the project is to compare the two, and see how they work at different crop stages.
The first reading taken is at the six-leaf stage. That’s an important because it’s the stage when the corn farmer is deciding whether to side-dress nitrogen to maximize yield potential.
The second reading offers better prediction, with the crop circle sensor, but that’s a little late, unless the farmer has had a catastrophic incident, like a 6-inch rain, which could have leached nitrogen out of the area where it is used.
The difference between Sharma’s work and what has been done before is that another factor — the height of the crop multiplied by the crop color readings — gives much clearer representation of the potential yield, Franzen says. The predictability gets two or three times better when the height data is added to the optical data.
“The way we see it working in the future is tying a crop height sensor along with these active optical sensors — the two of them together building algorithms that will let a farmer know if he needs supplemental N (nitrogen) as he goes through the field,” Franzen says.
“Scientists are developing the basic algorithms and, at the same time, we’ll be testing satellite image NDVI against our algorithms to see if they’re similar,” Franzen says.
“If they are, we can use the satellite images to predict which fields a farmer or an ag industry need to look at for supplemental nitrogen,” Franzen says. “If a person has 20 fields, does he really need to go through all 20 fields? Or are there some of them he can eliminate because they’re okay? He may have to visit only 10. That’s big logistic help for him."
On a larger scale, an ag industry that has 100 fields to deal with may use the technology as an indicator of how many of them will need supplemental nitrogen, and how much fertilizer to stockpile in advance.
Franzen sees the system being managed like this: “When the initial fertilizer is put on in the fall or pre-planted in the spring, the farmer will put some but not all of his nitrogen on the field because of the risk of loss in the first six weeks or so,” he says. “They’re planning on side-dressing, or top-dressing (nitrogen).
“Within that field, there’ll be a short strip — 100 feet or so, where they’ll double the nitrogen rate, maybe up to 200 pounds of N. That’ll serve as the field standard. It’ll be within the field, within the variety, so the variety differences of color don’t matter. The difference between the nitrogen-rich strip and the rest of the field tells you whether it needs supplementation or not. In our work, sometimes with a check plot with no added nitrogen yields just as much as 200 pounds of N, which means you didn’t need any nitrogen in the first place.
“There are things we can’t explain — the amount of nitrogen that comes from the soil naturally in some years that a farmer isn’t going to know. It varies from year to year. By putting that nitrogen-rich strip in the field, you have a standard within the field, within the variety, to say whether you’re going to need more nitrogen or not.”
Oklahoma State University developed technology that Trimble Navigation Ltd. of Sunnyvale, Calif., has rights to, Franzen says. A sensor is placed before the tractor, before the boom, or a few feet behind where the spray nozzles come out, and controller within a couple of feet of each other.
“In that brief amount of time, it’s doing the algorithms, doing the things that the computers do, and it’s turning off different nozzles to put on different rates of fertilizer into the ground,” Franzen says. “(The nitrogen) could be put on with a coulter, as a stream between the rows, or through the anhydrous tank.”
Crop height sensors are being developed in Germany and elsewhere, probably using a laser that would average the height along the way.
“Imagine an Excel spreadsheet with your NDVI, your light readings in one column and your heights right next to them, doing the algorithm rapidly as it goes through the field.”