Episode 4: Forecasting a Crop's Water Needs

Oct 22, 2013

We live in an ambitious desert society here in Utah. Food crops only grow here because we irrigate them with water we divert from rivers and reservoirs.

There's no chance that agriculture could flourish in Utah on rainfall alone. But one way to create more water in our system is to be more efficient with what we have; so researchers are working on making irrigation more predictable - think forecasting. Since the future of Utah's agricultural  industry might depend on the success of this research, we decided to learn more. Matt Jensen has the story.


Farmers of tomorrow will face a tricky problem in Utah. Statewide, the available land we can grow food on is limited, and the supply of water for irrigation - well that’s going to start shrinking, too.

Here’s the problem: across the Beehive State, and much of the world, urban populations are getting bigger, meaning thirsty cities will get more water, and farms – they’ll get less.

More than 80 percent of the water we divert in Utah goes to agriculture. That’s a lot of water, and experts say it’s not a very efficient use of a resource in such an arid state.

So how can we grow more food and have sufficient water for our urban centers? Engineers say the tradeoff has to come from better irrigation. We have to irrigate smarter, and researchers in Utah are trying to do exactly that.

At the Utah Water Research Laboratory in Logan, researchers are using satellite images of farmers’ fields to measure how much water a crop loses through evapotranspiration. That means the water a plant releases to the atmosphere and the water that evaporates from its soil.

This Crop Water Stress Map shows an area west of Delta, Utah. The colors indicate water needs.
This Crop Water Stress Map shows an area west of Delta, Utah. The colors indicate water needs.
Credit Utah Water Research Laboratory

Farmers have been trying to figure this out decades, using fancy equations or by just picking up a handful of dirt and taking a guess. But measuring evapotranspiration across an entire crop is difficult, meaning farmers often over-irrigate or under-irrigate. In fact globally, it’s estimated irrigation for agriculture is only 30 to 40 percent efficient.

Scientists at the water lab are developing better ways to more accurately measure evapotranspiration and creating new technology to forecast it - something that’s never been done before in irrigation engineering. Lab director Mac McKee:

"Actually what you're talking about is the next great important step in irrigation engineering," he said.

The satellites pictures capture information from multiple spectrums and, when combined with weather data from the farm, they can show how much water a crop loses pixel by pixel for the day the satellite, known as Landsat, was overhead.

"So you're getting a measure of the depth of water in each of those Landsat pixels that the plant plus soil gave up to the atmosphere on that day," says McKee. "The problem is you're only getting a measure of what happened, not what it will be in the future."

That’s where research engineer Alfonso Torres-Rua comes in.

"Forecasting evapotranspiration is relatively new," he said. "While the theory behind the evapotranspiration concept has been studied for several years, the forecasting part is the difficult part because it's related with the weather variables."

By adding those weather variables into the mix and some new computer modeling that’s been pioneered at the water lab, Torres-Rua says it’s possible to anticipate a crop’s water needs a week or so in advance.

The satellite images and weather information are loaded into an enormous database that’s studied for patterns. Torres-Rua and his team can then use the information to forecast evapotranspiration.

The implications are huge. Instead of watering an entire field uniformly, automated irrigation systems could deliver a specific dose of water based on the information from a single pixel on the satellite image.

Again, Mac McKee:

"This kind of technology could be used not only by irrigators but especially by managers of large, complicated irrigation systems to use water more efficiently," he said.

The satellites provide a resolution of about 30 by 30 meters per pixel. That’s a good start, but Torres-Rua wants to look much closer. Enter Aggie Air - an unmanned aerial vehicle with a love for water.

These small autonomous aircraft with sophisticated onboard cameras were developed by the water lab to take high-resolution pictures of irrigated agriculture. The imaging technology on board can deliver sub-meter resolution of the fields below, giving Torres-Rua the data he needs to develop tools for precision agriculture.

With this technology, Torres-Rua says farmers could get closer to 60 percent efficiency with irrigation water.

The water research lab has been using the system on an alfalfa crop near Delta, and will present the data at the Utah Water Users Association conference in March. Reporting for Five Billion Gallons, I’m Matt Jensen.

Five Billion Gallons is a production of Utah Public Radio, supported by Penn State Public Media's Think Outside the Pipes" radio initiative. We're part of "Water Blues, Green Solutions," a documentary film coming to public TV airwaves in January 2014.