Data-Driven Estimation of Actual Evapotranspiration to Support Irrigation Management

Antoine Richard, Offer Rozenstein, Lior Fine, Nitzan Malachy, Cédric Pradalier, Josef Tanny

Keywords: Precision agriculture and farm management, Remote sensing

TL;DR: Evaluation of data-driven, methods for support irrigation management in a real-world scenario.

Abstract:

Recent advances in remote sensing and machine learning show potential for improving irrigation efficiency.
In this study, two independent methods to determine the irrigation dose in processing tomatoes were tested in an irrigation experiment.
The first method used multispectral imagery acquired from an unmanned aerial aircraft (UAV) to estimate the FAO-56 crop coefficient Kc.
The second method used an Artificial Neural Network (ANN) to predict latent heat fluxes using meteorological variables from a nearby meteorological station.
An irrigation experiment was conducted, where the farmer was instructed through a mobile application with updated irrigation recommendations.
Both methods were compared against an expert guided control treatment.
Yields, water use efficiency, and Brix levels were measured, and showed to be on par with the control.
Additionally, both methods estimated ET at a near-perfect agreement with best-practice irrigation.
These results demonstrate the potential of machine learning techniques and aerial remote sensing to quantify crop water consumption and support irrigation management.