Calibration of CERES-Barley model using inverse modeling method under deficit irrigation conditions

Subject Areas : Farm water management with the aim of improving irrigation management indicators

Behnam Ababaei ¹ , Mahdi Sarai Tabrizi ² , Bahman Farhadi Bansouleh ³ , Teymour Sohrabi ⁴ , Farhad Mirzaei ⁵

1 - PhD, University of Tehran, Department of Irrigation and Reclamation Engineering, Karaj, Iran
2 - Young Researchers Club and elites, Science and Research Branch, Islamic Azad University, Tehran, Iran
3 - Assistant professor, Department of Water Engineering, Razi University, Tagh Bostan, Kermanshah, Iran
4 - Professor, University of Tehran, Department of Irrigation and Reclamation Engineering, Karaj, Iran
5 - Assistant professor, University of Tehran, Department of Irrigation and Reclamation Engineering, Karaj, Iran

Received: 2012-06-21 Accepted : 2012-12-20 Published : 2013-03-12

Keywords: Barley, Calibration, Deficit irrigation, Inverse Modeling,

Abstract :

Crop simulation models are used for understanding plants response to the change of surrounding environment and to evaluate crop physiological and morphological characteristics to enhance crop production and to contribute to the efficient use of water and soil resources. Yet, the evaluation of these models is a prerequisite for their use in assessing different management strategies. In this study, CERES-Barley from DSSAT software package (version 4) was calibrated by using inverse modeling method and PEST model. This study was conducted with 11 experimental treatments each with three replicates in the form of complete randomized block design. The irrigated treatments were consisted of 100 (T100), 90 (T90), 80 (T80), 70 (T70), 60 (T60), 50 (T50), 40 (T40), 30 (T30), 20 (T20), and 10 (T10) percent of crop water requirement. The rainfed treatment was consisted of one irrigation at planting (November 1st) and the other at first spring (April 8th) before introducing water stress. The results indicated that the model could reasonably well simulate these amounts up to T50 with acceptable accuracy. The calculated SRMSE statistics between T50 and T100 treatments for the simulated grain yield, straw yield, and total crop yield was 5.5, 10.9, and 3.6 percent, respectively. The corresponding R2 values were obtained to be 95, 69, and 93 percent, respectively.