The parameterization and evaluation of the APSIM model to simulation of yield and phenological traits of some wheat, barley and triticale cultivars (Case study: Alborz province)
Pages 151-168
https://doi.org/10.22059/jci.2026.406960.2964
Sorayya Navid, Mohammadreza Jahansouz, Saied Soufizadeh
Abstract Objective: This study aimed to parameterize and evaluate the APSIM-Wheat and APSIM-Barley sub-models for simulating the yield and phenological traits of wheat, barley, and triticale cultivars in Alborz Province, Iran.
Methods: Model inputs included soil, climatic, plant, and management data. A four-year experiment was conducted to collect the required information. For model parameterization and determination of genetic coefficients, two-year field experiments were performed using a randomized complete block design with 14 treatments (six barley, six wheat, and two triticale cultivars) and three replications—conducted at the Atomic Energy Organization farm (2014–2015) and the Faculty of Agriculture, University of Tehran (2016–2017). Genetic coefficients were identified from field data, and the model was locally calibrated. For model evaluation, farm sampling was carried out during the 2018–2019 and 2019–2020 growing seasons. Under real farming conditions (farmers' management), 30 barley and 30 wheat farms were selected across Alborz Province. A comprehensive questionnaire was used to collect information on farm history, planting, management, and harvesting operations, along with overall farm management practices. Soil and plant samples were also collected to assess crop growth status.
Findings: Simulation of flowering and physiological maturity stages using APSIM-Wheat and APSIM-Barley showed strong agreement between simulated and observed values across all wheat, barley, and triticale cultivars. The model predicted phenological traits with excellent quality and acceptable accuracy, with normalized root mean square error (nRMSE) values below 10%. In both experimental years, nRMSE values for grain and biological yield were below 5% for all cultivars. Additional evaluation metrics (CRM, D-index, and R²) further confirmed the robustness of the sub-models. Agreement between simulated and observed traits was higher for triticale than for wheat and barley. Among barley cultivars, yield simulations showed better agreement than those for wheat cultivars.
Conclusions: The estimated genetic coefficients and APSIM sub-models can be reliably used to predict phenological dates and yields of the studied cultivars across diverse regions and environmental conditions — including varying moisture regimes, fertilizer levels, and sowing dates — without the need for time-consuming and costly field experiments.















