Simulation-optimization based real-time irrigation scheduling: A human-machine interactive method enhanced by data assimilation

Efficient irrigation scheduling is crucial for both improving crop production and saving irrigation water use in arid/semi-arid agricultural regions threatened by water shortage and soil salinity. However, irrigation scheduling optimization is hindered by the uncertainties of data and optimization model, and adopting the optimal irrigation scheduling is subject to farmers' acceptance. To effectively tackle these challenges, this paper presents a novel human-machine interactive framework for real-time irrigation scheduling (RIS). The developed modeling framework couples a simulation-optimization model, irrigation farmers, and a data assimilation procedure within the human-machine interactive framework for RIS. The proposed approach is capable of: 1) searching optimal irrigation scheduling through the simulation-optimization model; 2) making actual irrigation decisions based on farmers' experiences, knowledge, behaviors, or optimal solutions; and 3) updating soil water content based on the model simulations and real-time observations at each time period. The RIS is applied to a real-world case in a typical arid agricultural region of China. Based on the comparisons with historical irrigation records and a tradition simulation-optimization model, the proposed RIS can not only achieve higher economic benefit with less irrigation water allocation quotas, but also improve irrigation efficiency. This study contributes to the methodology by integrating computer model, real-time observations and farmers' experiences to optimization modeling framework for supporting sustainable irrigation water management.

Automated summary

This paper presents a novel human-machine interactive framework for real-time irrigation scheduling, which can effectively tackle the challenges of data uncertainties and optimization models while considering farmers' acceptance. The framework searches for optimal irrigation scheduling through a simulation-optimization model, makes actual irrigation decisions based on farmers' experiences, and updates soil water content using real-time observations and model simulations. Applied to a real-world case in China's arid agricultural region, the proposed framework achieves higher economic benefit with less irrigation water allocation quotas and improves irrigation efficiency compared to historical records and traditional simulation-optimization models. This study contributes to sustainable irrigation water management by integrating computer models, real-time observations, and farmers' experiences into the optimization modeling framework.