Improving irrigation water management using agent technology

Abstract

Water use demand has increased steadily in Canterbury, New Zealand, over the last 20 years resulting in insufficient water availability. Prevailing irrigation management strategies in Canterbury consider soil moisture measurement up to a certain extent with no regard to crop’s development stages and sensitivity to water shortage; therefore, lacking to address the spatial and temporal variability of crop’s irrigation demand. This paper proposes an agent-based model that can be used to prioritise irrigation allocation to different crops on a farm. The proposed agent-based model takes into account the crop’s drought sensitivity, growth stage, soil type and the crop coefficient value when allocating water. It prioritises the crops based on their water requirements and generates an irrigation plan based on the predetermined water reduction. Our preliminary results showed that using this agent-based model, a significant water saving can be achieved even when water reduction is in effect, without impacting on production.