Putting sensors to work

Crop sensors can be useful to farmers, but it takes certain steps to make them work, says Robert Mullen, director of agronomy for Potash Corp. Three kinds of crop sensors are currently available, he notes: Crop Circle by Holland Scientific Inc., Ag Leader Technology; GreenSeeker by NTech Industries Inc., Trimble; and CropSpec by Topcon Positioning Systems.

“The three are basically similar, but the approach is different,” he says. “The real question is, how do they go about turning a sensor reading into a decision?”

The sensors convey the “health” of the plant. They work by sending light beams to the plant surface. Some light is absorbed and some is reflected. The sensor reads the light reflected by the plant. Essentially, a plant with a paler shade of green shows up as less healthy than a plant with a deep-green color.

However, without additional information, how does an operator identify the cause of poor health? “Can we assume it is always nitrogen?” Mullen asks.

This is a specific problem that calls for a specific solution, he says. Most researchers recommend a grower plant a nitrogen-rich strip. This is used to calibrate the machine to the color of plant that, without doubt, has plenty of N.

“You need a zeroing reference,” he says. “In the absence of a reference strip, you may get statistics that are OK, but a reference strip is better.”

It is also critical to apply the N at a time when the plant can use it. V-7 to V-8 are points where N is being accumulated in the plant, as you can see from the chart. Since the corn crop is likely to be more than waist-high at this point, a high-clearance applicator is necessary.

Mullen also recommends the unit be loaded with an algorithm that is suitable for the area in which it is being used. Local weather patterns are different, and algorithms should reflect typical conditions.

“The technology is reliable,” he says. “It’s not really the sensors that matter. It’s the calibration of the data or the algorithm that determines the usefulness of a unit.”

Algorithms or equations have been designed for different areas. Mullen recommends using an algorithm he created for Ohio while he was the fertility specialist at the Ohio State University. Among other things, this algorithm takes into account a greater likelihood of N being lost, he says.

“Ohio’s response early in the season is more likely to underestimate N response at harvest due to weather patterns after sensing,” he says. “That’s why Ohio requires a different fertilizer decision algorithm than one developed in the western Corn Belt.”

ScreenHunter_01 Jan. 07 14.01.jpg

This article published in the February, 2012 edition of OHIO FARMER.

All rights reserved. Copyright Farm Progress Cos. 2012.