Irrigatmo: no-moving parts system for feed-back and feed-forward irrigation scheduling

 

Abstract 

The weather-based approach quantifies the crop water requirements (CWR) using a simplified agrohydrological model coupled with meteorological sensors.  The FAO56 model (Allen et al., 1998) is one of the most used bucket models for CWR. In this model, the daily ET0 is usually estimated by the FAO-Penman-Monteith (PM), which needs as inputs standard atmosphere forcings acquired from weather stations, that often are equipped with ordinary mechatronics sensors that require regular maintenance. An atmometer (ETgage) is an accurate sensor with no moving parts that continuously measures the ET0 based on a physical analogy of the crop reference.  

This study aims to design and validate an expert system, named Irrigatmo, to manage irrigation based on the combined application of the feedforward- (FFc) and feedback- (FBc) control irrigation scheduling protocols. The FFc protocol comprises a Kc-based mass balance model with a modified atmometer and FDR sensors for sub-hourly ET0 and soil water content (SWC) measurements. At the same time, the FBc protocol uses the SWC to quantify the critical condition and the crop stress coefficient to adjust the Kcb value used in the bucket model. The system was implemented in proprietary logic (CR300, Campbell Scientific Inc.) and open-source logic (Arduino Mega 2560, Arduino). The core of the system implements a weather-based water balance model, trained by a modified atmometer and soil moisture sensor for sub-hourly scale ET0 and SWC, as well as an infrared thermometer and a contact thermocouple for quantifying the crop water stress index (CWSI). The ETgage was modified by integrating a pressure transducer sensor, calibrated to measure the water level inside the atmometer tank continuously.  

The results showed that Irrigatmo accurately and rapidly detected the changes in atmospheric and soil water conditions. The system can directly calculate the evapotranspiration reduction factor (Ks), estimating the CWSI based on canopy temperature measurements. This could overcome the uncertainty in the models associated with the water stress function based solely on the soil moisture. The system was built and calibrated within the AgrHySMo laboratory of DiSAAA-a and validated on a commercial kiwifruit orchard of Actinidia chinensis var. chinensis 'Zesy002'. The field testing made it possible to validate the system's ability to model the water stress functions of the crop and the sensitivity to identify the critical water status conditions that mark the transition to a limiting condition. Irrigatmo could manage irrigation autonomously, activating or turning off the solenoid valves, and returning to our field the amount of water lost during the evapotranspiration processes. Future perspectives consider the implementation of the proposed system in a wireless sensor network (WSN) and at the interfacing of the WSN nodes with aerial platforms where the edge-computing systems specialized also in the control of IoT-irrigation actuators will be located.