Crop-growth-based spatially-distributed optimization model for irrigation water resource management under uncertainties and future climate change

The spatially distributed AquaCrop-multiobjective robust possibilistic programming (distributed AquaCrop-MORPP) model that considers spatially-distributed land, soil and management features has been proposed for irrigation scheduling optimization under uncertain conditions. Compared with traditional simulation-optimization models, it improved the robustness of decision making by optimizing irrigation amounts and dates considering physical crop growth process and the spatial heterogeneity of soil, crop and irrigation. Additionally, it enhanced the AquaCrop-optimization model by handling uncertainties presented as fuzzy and stochastic variables. Moreover, it enhanced the applicability of multi-objective programming (MOP) by balancing contradictory relationships between the objectives of the net economic benefit and field water use efficiency, which were simulated and calculated by the spatially-distributed AquaCrop model. Fifty groups of optimal Pareto solutions were obtained by solving the spatially-distributed AquaCrop-MORPP model using a fast and elitist multi-objective genetic algorithm (NSGA-II) method. Subjective decision-making methods (e.g., Technique for order preference by similarity to an ideal solution (TOPSIS) and prospect theory) were used to select specific alternatives from the Pareto solutions to provide recommendations for managers from different viewpoints. The alternatives that maximized field water use efficiency or net economic benefit separately were selected to cope with emergencies. To obtain optimal irrigation scheduling under future climate change, the RCP4.5 future climate scenario was integrated with the spatially-distributed AquaCrop-MORPP model. The model was verified by applying it to the Yingke Irrigation District (YID), Heihe River Basin (HRB), China. The results showed that the maximum field water use efficiency and net economic benefit in 2021 improved by 24% and 1.3% than those in 2012 for the alternative selected by TOPSIS method. The irrigation scheduling with varied irrigation dates improved by 3.7% and 5% than the irrigation scheduling with fixed irrigation dates for field water use efficiency and net economic benefit, separately. The study provides a framework about how to build spatially-distributed crop simulation-optimization model with consideration of tradeoffs amid multiple objectives to optimize irrigation schedules, and how to select alternatives based on managers’ risk attitudes, and offer a series of Pareto solutions for managers.