Development and simulation of sensor-based irrigation control strategies for cotton using the VARIwise simulation framework

• Two site-specific sensor-based irrigation strategies were simulated in VARIwise.
• Iterative Learning Control (ILC) produced highest yield with soil–water data input.
• Iterative Hill Climbing Control (IHCC) performed best with soil-and-plant data input.
• Both sensor-based strategies were superior to the industry-standard strategy.

Feedback control systems offer opportunities to accommodate spatial and temporal differences in crop water requirement and to improve the automated irrigation of field crops via real-time data from in-field plant, soil–water and evaporation sensing. This paper describes two sensor-based strategies applied to irrigation control, ‘Iterative Learning Control’ (ILC) and custom-designed ‘Iterative Hill Climbing Control’ (IHCC), implemented in the control simulation and evaluation framework ‘VARIwise’. Simulation of an irrigated cotton crop using soils and merged 1999–2004 weather data of SE Queensland, Australia, and represented by the performance of the well-validated cotton growth and production model OZCOT, permitted the relative performance of differing sensor data types and availability to be evaluated (both as alternatives and in combination) in meeting the requirement to optimise either crop yield or water use efficiency. These simulations indicated that ILC would perform better at maintaining soil–water deficit, whilst IHCC would be better at maximising crop yield when plant and soil sensors were utilised in combination. This work demonstrates that the optimal choice of field sensor(s) and control strategy will be a function of the irrigation objective and the spatial and temporal availability and type of field measurements.