Available online: http://scholarsmepub.com/ 356 Haya: The Saudi Journal of Life Sciences (SJLS) ISSN 2415-623X (Print) Scholars Middle East Publishers ISSN 2415-6221 (Online) Dubai, United Arab Emirates Website: http://scholarsmepub.com/ Calibration and Validation of DSSAT Model for Simulating Wheat Yield in Bangladesh Apurba K. Choudhury *1 , S. Ishtiaque 1 , R. Sen 1 , M.A.H.S. Jahan 1 , S. Akhter 1 , F. Ahmed 1 , Jatish C. Biswas 2 , M. Manirruzaman 2 , M. B. Hossain 2 , M. Muinnuddin Miah 3 , M. M. Rahman 3 , Taslima Zahan 1 , A.S.M.M.R. Khan 1 and Naveen Kalra 4 1 Bangladesh Agricultural Research Institute, Gazipur, 1701, Bangladesh 2 Bangladesh Rice Research Institute, Gazipur, 1701, Bangladesh 3 BSMRAU, Salna, Gazipur 1706, Bangladesh 4 Division of Agricultural Physics, IARI, New Delhi, 110012, India Original Research Article *Corresponding author Apurba K. Choudhury Article History Received: 10.03.2018 Accepted: 22.03.2018 Published: 30.04.2018 DOI: 10.21276/haya.2018.3.4.3 Abstract: Crop simulation models are used in predicting crop productivity under various crop management options and changing climatic parameters that require determination of genetic coefficient of a crop cultivar. Successful use of a crop model depends on the accuracy of calibration and validation of different parameters. This paper aimed to evaluate the DSSAT (v4.6) crop model in Bangladesh for wheat production. Genetic co-efficient of four wheat cultivars (BARI Gom-25, 26, 27 & 28) were determined through DSSAT model. Experimental data on irrigation and nitrogen effect on wheat production was used for model calibration and date of sowing effect for model validation. Model evaluation results showed closer estimation of crop growth duration, grain and biomass yields. Percent error difference in grain yield of wheat varieties (BARI Gom- 25, 26, 27 and 28) between simulated and observed values were 10.98%, 8.70%, 10.79% and 8.94%, respectively. Relationship between simulated and observed grain yields, and simulated and observed crop duration both at calibration and validation process are strong having higher R 2 value. The model has been successfully calibrated and validated for wheat growing in Bangladesh environment and can now it can be taken for further applications in natural resources management and climate change impact studies. Keywords: Calibration, Validation, DSSAT Model, Wheat, Bangladesh. INTRODUCTION Wheat is the second most important cereal crop in Bangladesh [1]. It is very much sensitive with climatic parameter specially temperature. Though it is cultivated in winter, the season is very short. So, wheat crop suffer from higher temperature at the later part of crop growth and development. Yield also depends on various management options like date of sowing and irrigation, etc. Bangladesh Agricultural Research Institute developed 30 wheat cultivars. Of them, BARI Gom-25, 26, 27 and 28 are latest and tolerant to different stress conditions [2]. Crop models are effective tools to predict crop productivity under different management options and climatic conditions. In the contest of climate change, importance of crop model in simulating crop production under different climatic scenario was increasing day by day [3, 4]. Bangladesh agriculture is highly vulnerable in the contest of climate change. About 4.1% wheat grain yield will reduced if 1 o C temperature rise [2]. Successful use of crop model depends on proper calibration of models. Determination of genetic coefficient of a cultivar can be obtained from proper model calibration. Calibrated crop models with cultivar parameters can be used to optimize crop management [5] to evaluate the impacts of climate change [6] or to develop options and to optimize resource use [7]. DSSAT is a popular crop model used over 100 countries for more than 20 years [6]. It is a microcomputer software package, that provides a shell program for the interface of crop-soil simulation models, data for soil and weather, and programs for evaluating management strategies. DSSAT includes more than 40 crop growth model. Among them, CERES-Wheat is most widely used crop simulation model. Crop modelling study especially DSSAT on different crops is meagre in Bangladesh. Cultivar coefficients of popular wheat varieties of Bangladesh are not included in the cultivar database of DSSAT [8]. Therefore this study was undertaken: i) to generate genetic co-efficient of wheat cultivars required for