Aqua-MC as a simple open access code for uncountable runs of AquaCrop.

Understanding uncertainty in crop modeling is essential for improving prediction accuracy and decision-making in agricultural management. Monte Carlo simulations are widely used for uncertainty and sensitivity analysis, but their application to closed-source models like AquaCrop presents significant challenges due to the lack of direct access to source code. This study introduces Aqua-MC, an automated framework designed to facilitate Monte Carlo simulations in AquaCrop by integrating probabilistic parameter selection, iterative execution, and uncertainty quantification within a structured workflow. To demonstrate its effectiveness, Aqua-MC was applied to wheat yield modeling in Qazvin, Iran, where parameter uncertainty was assessed using 3000 Monte Carlo simulations. The DYNIA (Dynamic Identifiability Analysis) method was employed to evaluate the time-dependent sensitivity of 47 model parameters, providing insights into the temporal evolution of parameter influence. The results revealed that soil evaporation and yield predictions exhibited the highest uncertainty, while transpiration and biomass outputs were more stable. The study also highlighted that many parameters had low impact, suggesting that reducing the number of free parameters could enhance model efficiency. Despite its advantages, Aqua-MC has some limitations, including its computational intensity and reliance on the GLUE method, which may overestimate uncertainty bounds. To improve applicability, future research should focus on parallel computing, cloud-based execution, integration with machine learning techniques, and expanding Aqua-MC to multi-crop studies. By overcoming the limitations of closed-source models, Aqua-MC provides a scalable and efficient solution for performing large-scale uncertainty analysis in crop modeling.