Researchers create a new irrigation optimization tool.

An irrigation optimization tool developed by Stanford researchers has the potential to assist farmers in reducing water consumption.

The “evapotranspiration” process, which involves the evaporation of water into the atmosphere and the uptake of water by plants, is quickly estimated by the tool. The new Stanford modeling tool is 100 times faster than the most recent methods for estimating such evapotranspiration while maintaining high levels of accuracy.

In practice, the tool may significantly shorten the amount of time required to create strategic, effective irrigation schedules that position watering and sensing equipment in the most effective locations across entire farms. The tool could even crunch data quickly enough on a smaller scale, field by field, to adjust irrigation as the weather changes almost immediately.

“With this research, we’re helping to deliver on the promise of smart agriculture, which is to feed billions of people around the world while also preserving our planet for future generations.”

Daniel Tartakovsky, a professor of energy science and engineering .

Weiyu Li, a Ph.D. candidate in energy science and engineering and the lead author of a study in Water Resources Research describing the findings, stated, “Evapotranspiration is a critical piece of information for designing efficient irrigation systems.” Li is the first and only recipient of a Siebel Scholarship at the Stanford Doerr School of Sustainability, and she is a member of the 2023 class.

Overall, the research is a step forward for smart agriculture, which increases crop yields while conserving natural resources by making use of cutting-edge technologies and methods like big data and the Internet of Things.

Senior study author Daniel Tartakovsky, an energy science and engineering professor and Li’s advisor, stated, “With this study, we’re helping to deliver on the promise of smart agriculture to continue sustainably feeding billions of people worldwide and preserving our planet for future generations.”

Straightforward upward, complex level
The customary way of representing evapotranspiration has depended on what scientists call the upward stream supposition. The water applied during irrigation is modeled in this way as only going straight into the soil. The fact that the water can and does flow horizontally is not taken into consideration. The vertical-flow assumption has been used as a kind of computational shortcut because smart agriculture needs to process a lot of data. Li stated that while the method is adequate for some irrigation modeling requirements, it has significant room for improvement in terms of its output.

Li explained that the vertical-flow assumption is insufficient for truly intelligent agriculture, particularly when it comes to “drip irrigation.” Drip irrigation, as the name suggests, involves delivering water precisely and slowly to the root zones of plants so that it can be absorbed with the least amount of evaporative loss. Drip irrigation is mostly used in arid areas, like a lot of California, where traditional methods of watering fields by inundating them use a lot of water.

Additionally, smart agricultural systems optimize timing. They water a plant just when it’s required, contingent upon elements like climate and the plant’s phase of development. ” “Tartakovsky stated that irrigation has historically been largely separated from the needs of the plant at any given time.” That is not the case with smart agriculture-based drip irrigation.

Part of the test of shrewd agriculture, then, is knowing where to best position dampness sensors and drippers. This tool aims to provide that guidance based on real-world and nearly real-time conditions, as opposed to the approximations and assumptions that are used in existing designs.

Better algorithms

Li and Tartakovsky used algorithms to improve data crunching and produce high-quality results while developing the tool. The researchers combined the enhanced Kalman filter and maximum likelihood estimation algorithms for the new study. Predictions based on measured data are used by the algorithms to reduce uncertainties based on subsequent measurements.

“We plug genuine information estimations of soil dampness and root water take-up into our model, which works on how we might interpret the generally actual framework and the calculation’s presentation,” Li said. “Our review is quick to consolidate this sort of algorithmic methodology and apply it to trickle water systems.”

The Stanford researchers simulated a 5-by-33-foot plot of land, roughly the size of a short row of planted crops, to see how accurate and effective their method was.

A precise estimate of the evapotranspiration rate for the test plot of land could be calculated in less than ten minutes using the new modeling tool. According to other recent studies, the computational time would have been approximately 1,000 minutes longer if only an enhanced Kalman filter had been used. This portion of time, which amounts to nearly 17 hours, cannot be utilized for timely, smart agriculture. An irrigation optimization system based on our modeling tool, on the other hand, could respond to changing conditions almost immediately,” Li stated.

The computational time needed to achieve the goal of optimizing the initial design of drip irrigation systems for an entire farm, which can cover thousands of acres, becomes almost prohibitive. “When faced with major installation projects, you can start to see why irrigation system designers have relied on the simplified vertical-flow method,” Li stated.

Looking forward, the Stanford scientists intend to perceive how well their demonstrating device functions in true settings when sent on a functioning homestead. “We next need to play out a ‘field’ test, in a real sense,” Tartakovsky said. “With all the variables presented by real crops, real drippers, real sensors, and real weather, we look forward to further fine-tuning our strategy.”

More information: Weiyu Li et al, Fast and Accurate Estimation of Evapotranspiration for Smart Agriculture, Water Resources Research (2023). DOI: 10.1029/2023WR034535

Topic : Article