Evaluation of AquaCrop Model in Simulating Wheat Yield in Four Consecutive Years in Special Sites of Khuzestan Province, Iran

Bavi, Adel

doi:10.22077/jdcr.2026.9191.1134

Evaluation of AquaCrop Model in Simulating Wheat Yield in Four Consecutive Years in Special Sites of Khuzestan Province, Iran

Articles in Press

Document Type : Original Article

Author

Adel Bavi

Assistant Professor

https://doi.org/10.22077/jdcr.2026.9191.1134

Abstract

Although numerous studies have employed the AquaCrop model to simulate wheat growth, the evaluation of this model over multiple years and on farmers' fields has received limited attention from researchers. Given Iran’s agricultural policies aimed at increasing wheat yields, it is crucial to assess the model's performance under real-field conditions over several years. To this end, data were collected from wheat farms spanning 4,000 hectares in the cities of Shush, Ahvaz, Hoveyzeh, and Khorramshahr during 2020 to 2023, in line with the national policy objectives. The model’s performance was evaluated using metrics such as Mean Bias Error (MBE), Root Mean Square Error (RMSE), Normalized RMSE (NRMSE), Agreement Index (d), Model Efficiency (EF), and Coefficient of Determination (R²). Results indicated that AquaCrop performed with excellent accuracy across all locations, with NRMSE values below 0.1. Over the study period, model accuracy improved, with NRMSE decreasing from 0.09 in 2020 to 0.05 in 2023. In all studied cities, the model tended to underestimate yields, as reflected by negative MBE values, a trend consistent throughout the years. Overall, the findings suggest that geographical location had little impact on the model’s accuracy. Instead, improvements over time, likely due to enhanced wheat yields on farms, contributed to increased model precision.

Keywords

Crop Modeling

Yield Trend

Wheat Yield

Wheat-based Farms

Subjects