For the past 10 years, CropMetrics has enabled irrigators to make more informed decisions based on soil moisture data feeding into the CropMetrics platform. Until recently, those decisions were made from real, physical probes installed in the field.
In 2019, however, CropMetrics has launched a new tool using a sensorless approach. The Virtual Predictor tool, launched earlier in the summer, uses a "virtual probe" that relies on modeling tools combined with different data streams and agronomist insights to make decisions.
In addition to irrigation recommendations based on the current soil moisture status, the Virtual Predictor tool also provides forecasts seven days out, allowing growers to plan ahead on a week-by-week basis.
“We set up a system that automatically feeds information on crop and soil conditions, weather and evapotranspiration into a powerful data science model that allows us to make a soil water status forecast,” explains John Gates, chief science officer at CropMetrics. “It gives you an estimate of your soil moisture and soil water status over the next seven days. Your agronomist has access to that and leverages it in the recommendations that they write.”
The Virtual Predictor tool was launched as a limited release for 2019, but will be fully commercially available in 2020 as part of the CropMetrics core product offering. Pricing information will be announced before the end of the year.
Developing the tool involved testing in different environments and weather conditions — and working closely with growers and agronomists to ensure the system's accuracy.
FIELD COMPARISON: Developing the Virtual Predictor tool involved testing in different environments and weather conditions — and working closely with growers and agronomists to ensure the system's accuracy.
One of the big advantages of the Virtual Predictor tool is ease of use, Gates says. The platform interface was designed to match growers' decision-making processes.
"If I've got a really busy day ahead, what about my irrigation management plan, if anything, needs my attention first today?" Gates says. "What are my problem spots? What are my watch-outs? The first thing you see when you log in is a screen of potential watch-outs. If there are fields or locations where we've got problem areas with soil water conditions, that's the first thing you see. We flag all that for you. From there, you can click into more detailed views of anything you need at that point."
Of course, many users opt to let their agronomist give them a recommendation. All fields enrolled in a CropMetrics product are assigned a certified CropMetrics agent that keeps an eye on the field data. The agronomist or agent has their own account connected to the grower account, and will be able to provide updates and recommendations.
"Growers are sometimes a little skeptical of prediction models in agriculture, and we appreciate that since the industry overall has seen various levels of success with them in the past," says Nick Lammers, agronomic services lead at CropMetrics. "But with the Virtual Predictor, there's a set of human eyes on every one of these fields. So, although the model is projecting where it thinks you're going to be in terms of soil moisture, there's still a human factor involved."
The Virtual Predictor tool uses radar-interpolated rainfall data. However, in some cases, rainfall data may need to be entered manually or measured more directly via a soil moisture probe or rain gauge.
"We do have fields that have anomalies in them," Lammers says. "Like a high-water table, for instance. The model doesn't know you have a high-water table, so it's not going to work as well there. When you have one-off situations, you're going to want to fall back on technology that will make direct measurements.
"On flat, well-drained fields that don't have any anomalies, the model will work very well. If you take that data and add it to the data points you're bringing in from probes, you can enhance the model and make even more accurate projections."
Along with interpolated rainfall data and physical soil characteristics, the prediction system also needs to estimate the crop's condition — including response to temperature, accounting for crop type, variety, planting date, maturity, growing degree days and rooting depth.
In addition, CropMetrics uses an uncommon but accurate approach to calculate evapotranspiration, inferring it from a detailed soil moisture time series using data from soil moisture probes.
"Really, no one of these individual science pieces is super new," Gates says. "I think the big step forward is integrating all those things into a system that doesn't just rely on one type of data. We’re integrating the whole system to use any data stream that has value and can bring an advantage to our agronomist and growers. That integration is where a lot of the power comes from.”
Moving forward, Gates says CropMetrics plans to work with partnering companies to bring additional data streams to the platform.
"I think we're ending this era of data silos between company A and company B, or input A and input B, and moving into a world where data can flow more easily according to the farmers' needs," Gates says. "This way, each one of your decision platforms is no longer as isolated, and the analytical models that CropMetrics runs can benefit from data on another input."
For 2019, CropMetrics also launched its mobile app — allowing users to access the CropMetrics platform from their iPhone or Android device.
"We've prided ourselves in the past about the power and the flexibility of our platform, especially around giving growers and agronomists the opportunity to write custom variable-rate irrigation prescriptions," Gates says. "Ease of use is something we really wanted to drive home, so we made sure this next generation of technology was powerful on-the-go, and that it also was more understandable and usable by busy growers."
"One of the things that we built into that approach was the agronomist irrigation recommendation, which comes along with our service package for every product tier we sell," Gates adds. "It's front and center in that app, so the first thing a farmer sees when they log on to CropMetrics is that communication with their trusted adviser, who has looked at the same data that they have access to and has helped them distill it down to a takeaway or an action."