Farmers may be getting tired of hearing about data, or how big data will make a difference. Yet the march toward putting massive information to use for better decision-making continues. The latest milestone is news that WinField United has launched an advanced crop model it's calling the Field Forecasting Tool.
An enhancement to the company's R7 tool, this new system models crop production for water, nitrogen and potassium use. It's the first crop model to manage three inputs and help with in-season decision-making on this level. Other crop models on the market look at nitrogen use alone.
"We have insights from the company Answer Plots with a lot of data we could use," explains Joel Wipperfurth, ag technology applications lead, WinField United. "We worked with a French company —ITK — to build an in-season decision-making tool."
The WinField United Answer Plot program has been going on for 19 years and has collected a wide range of data. In fact, the cooperative has amassed more than 6 million data points on hybrid and variety performance under a range of conditions. Innovations from those plots include measurement of corn hybrid response-to-nitrogen measurements, as an example.
But a pile of data does little good unless you can start using it to make decisions, and the FFT does just that. With ITK's help, WinField United has a tool that pulls in current-year information from satellite data, tissue testing and weather information to gauge crop performance. The model can also predict how that corn would yield on that specific field, given the hybrid or variety selected.
"We want to help the farmer achieve optimum yield," Wipperfurth says. "With satellite imagery and tissue testing, we're using the plant as the sensor in the field." Models are built on actual performance data, but then, given knowledge of specific parameters like soil type, nitrogen availability, previous-year crop and previous-year yield, it's possible to look ahead.
ITK and WinField United are using advanced machine learning tools to constantly upgrade and update the model in-season, too. As conditions change, the measurements can change over time, providing increased accuracy, Wipperfurth explains.
IN-SEASON VIEW: This image from the Field Forecasting Tool shows crop progress. The dark green area at the bottom is the actual crop, while the lighter green is actual nitrogen needed for maximum yield (note that the crop will be nitrogen-stressed). (WinField United)
Modeling the crop
For farmers that use FFT, the first step is to enter some key information, from tillage type to the hybrid planted. Previous-year information is also valuable — especially for nitrogen management. The upfront information is needed to match your practices to what is known about crop response under a variety of conditions.
"We had a pilot program of 22 owners with 2,700 fields enrolled to evaluate this as a new product," Wipperfurth says. "We've been working to develop the tool with ITK since 2012."
He adds that modeling in a computer is not unlike the knowledge an agronomist has in his or her head. It's valuable information about understanding local conditions, and how soils interact with crop management practices. But as new agronomists enter an area, they may not have time to gain all that knowledge, as veteran agronomists have. This system helps overcome that.
Mary Pat Sass, an ag technology specialist working in Wisconsin and northern Illinois, demonstrated the model for media recently. She explained that the system provides a current status for all fields, with a look at water, nitrogen and potassium availability matched to the current growth stage.
Sass showed results for a Wisconsin field, where weekly tissue tests have been taken to help monitor the accuracy of the model. The field she showed had a 367-bushel yield potential; it is an irrigated field in southern Wisconsin.
During the presentation, the model showed the field would hit 284 bushels per acre, given some early-season heavy rain events, and early nitrogen stress — less nitrogen available to the plant when needed. The model clearly showed that an application of nitrogen later in the season could help boost yield. "We look at multiple scenarios," she said. "We see areas where we can push up the yield. The model shows us opportunities."
The model currently works for corn and soybeans, but information for wheat is being evaluated for release in the future. "The model currently works in wheat, but it has not been calibrated with high-quality data," adds Wipperfurth. "We want to calibrate the model for wheat and have confidence in the data."
For more information about the Field Forecasting Tool, visit winfieldunited.com.
Crop model teaches lesson
When looking at a crop model for corn, there's one glaring message that stands out: The crop needs more nitrogen in later life cycles that most farmers may be providing. The second conclusion? Fall application of nitrogen may not have the following-year benefit many hope to achieve.
Jay Culver, technical agronomist, South Dakota Wheat Growers, works with farmers raising more than wheat. During a demonstration of the WinField United Field Forecasting Tool, he observed that nitrogen use for corn — as an example — does currently match how farmers are putting it on.
"What we're seeing is that splitting that nitrogen application is boosting efficiency," he says.
During a demonstration at a recent national Answer Plot tour event for WinField United, corn that got a spring application of 75 pounds of nitrogen, followed by a midseason 75 pounds, was outperforming fields that got 300 pounds in one shot ahead of planting. That's a 150 pound savings per acre in nitrogen applied.
Crop modeling tools like the Field Forecasting Tool can further demonstrate ways to boost nitrogen efficiency.