29 May 2023

Algorithms drive irrigation in smart agriculture


In an attempt to make irrigation more efficient, Stanford University has developed a new and more accurate system for calculating evapotranspiration. Measurement is shortened by 100 times

by Matteo Cavallito


Optimizing irrigation, that is, not wasting water through real-time knowledge of plant and soil hydration levels. That is the goal set by researchers at Stanford University in California, who have designed an innovative tool for reducing consumption. The technology, explains a statement from the university, quickly estimates water loss from the soil due to evapotranspiration, or the evaporation of water into the atmosphere and its absorption by plants.

According to the designers, the new tool allows the information to be processed “100 times faster while maintaining high levels of accuracy.” The idea is to drastically reduce the time it takes to process irrigation schedules by placing equipment in the best possible way. In other words, an example of smart agriculture, that is, the set of practices that aim, for the same yield, to reduce waste in crops through the use of technology.

The importance of “horizontal” data

The assessment of evapotranspiration, the researchers point out, is based on the so-called vertical flow assumption. This operation, in other words, takes into account only top-down water movement while ignoring horizontal flows with the aim of simplifying calculations. “While adequate in some large-scale applications,” says a study by Stanford University, “this assumption fails when the horizontal component of the local flow velocity is not negligible due to, for example, soil heterogeneity or drip irrigation.”

The need to measure and describe flows in each direction is essential for the development of anti-waste practices beginning with drip irrigation, which allows water to be distributed slowly and precisely to plant roots, thus limiting evaporation. This technique is mainly used in arid regions where water consumption is typically higher.

Algorithms are crucial

The new measurement tool is based on the combination of two different algorithms known as enhanced Kalman filter and maximum likelihood estimation. “We plug real data measurements of soil moisture and root water uptake into our model, which improves our understanding of the overall physical system and the algorithm’s performance,” explained Weiyu Li, researcher and lead author of the research.

“Our study is the first to combine this kind of algorithmic approach and apply it to drip irrigation.”

The simulation, conducted on a plot of about 3 by 3 meters, yielded excellent results. The precise calculation of the evapotranspiration rate took only about ten minutes. About one hundredth of the time it would have taken -almost 17 hours – if only the enhanced Kalman algorithm had been applied, the authors explained.

Reducing waste in irrigation crucial

Timely collection of information enables proper deployment of moisture sensors and drippers. The researchers’ goal is to test the system on farms to enable them to optimize irrigation. This is something that is becoming increasingly important, as the FAO has mentioned in the past.

Indeed, the UN organization estimates that at the end of the 20th century agriculture was responsible for an average of 70 percent of global water withdrawals. Without adequate corrections these will increase by another 14 percent by 2030.

“The interconnected systems of land, soil and water are stretched to the limit with agricultural systems at breaking point,” Feras Ziadat, coordinator of the 2021 edition of The state of the world’s land and water resources for food and agriculture report produced by FAO, explained to Re Soil Foundation. Addressing drought and water scarcity therefore requires “the adoption of new technologies and management approaches.”