Simulation of Yield, Evapotranspiration, Water Requirement and Water Use Efficiency of Wheat Using CERES-WHEAT-DSSAT Model in Shahrekord Plain

Statistical models and numerical simulations have been widely used to detect relationships between the climate and crops. However, the influence of non-climatic factors (such as cultivar and fertilizer changes on yield crop needs to be eliminated. For this reason, dynamic crop models include the SUCROS, Erosion Productivity Impact Calculator(EPIC), WOrld FOod STudies (WOFOST), Agricultural Production Systems sIMulator (APSIM), and Decision Support System for Agrotechnology Transfer (DSSAT) have been used in water, nitrogen and weather responses. Among these models DSSAT contains separate models for different crops and can quantitatively predict the growth and production of the annual field crops.

In this study, the data of Shahrekord Agricultural Meteorological Station and the data of Lysimeter Station were used to evaluate the correlation between the research results and Lysimeter data from Pearson correlation coefficient, and the RMSE, MAD and MSE are applied in order to calculate the error.

Lysimeter: The wheat evapotranspiration amount from the planting (20 of Octobers) to the harvest time (14 of July) is recorded as 611.24 mm. Precipitation during the winter is low but continuous and it is 127 mm that equivalent to the evapotranspiration at this time of growth. In the warm season, a quarter of the evapotranspiration is provided by rainfall. The average of winter evapotranspiration is 0.87 mm per day and in the growth season is 4 mm per day. Also from planting to harvest is 2.42 mm per day that is recorded its maximum 7.8 mm and its minimum 2.32 mm per day. The total amount of drained water during the growth is 76.04 mm that 8.8% of the total rainfall. It indicates that drainage water from the soil is low and irrigation has a high efficiency. CERES-WHEAT  Wheat evapotranspiration amount during the growth period is 413.51 mm by FAO Penman-Monteith and 489.53 mm by Priestley-Taylor. Precipitation during the winter is low but continuous and it is 127 mm that equivalent to the evapotranspiration at this time of growth. In the warm season, a quarter of the evapotranspiration volume is provided through rainfall. The average of winter evapotranspiration based on the F.P.M and P.T methods are 0.86 and 1.23 mm/day and in the growing season 2.98 and 3.11 mm/day, respectively. During the experiment, the evapotranspiration average is 1.59 mm/day for the FPM method that the maximum is 6.61, and the minimum is 0.379 mm/day. This amount is 1.88 mm/day for P.T method which the maximum is 5.64 and the minimum is 0.45 mm per day. The total amount of drained water during the growing period is 106.3 mm, based on the F.P.M method and 90.2 mm based on the P.T method.The correlation between farm data and the data obtained through the F.P.M method of CERES-Wheat model is 0.97, which for the P.T method is 0.92. The MAD, MSE and RMSE values obtained between the F.P.M method and farm data are 0.95, 0.95 and 1.57, respectively, and for the P.T method, 0.97, 1.47 and 1.21, respectively. With respect to correlation and MAD, MSE and RMSE value, it is found that the model is highly capable in simulating evapotranspiration and crop performance. Among the methods applied in determining evapotranspiration, the F.P.M method with high correlation and lower error value is more accurate than the P.T method.Water Factor: From the day 177 to 216 is considered the most sensitive stage of plant growth. Based on DSSAT output over a 25-day period (196 to 216 days) the water available is severely depleted and the plant may experience drought stress. At this stage of the growth, water deficiency should be offset by increasing the time and the amount of irrigation.Day 210 is the beginning time of the increase in evapotranspiration of the plant. During this period, the amount of water which is uptake from the soil was less than 1 time the plant demand. This period of stress was based on the FAO Penman- Monteith method between the 203rd and 210th days. During this period, the plant goes through its clustering and flowering stages, and water stress at this stage causes the growth of wrinkled and lean grain, resulting in reduced grain weight and reduced crop yield. Water scarcity must be compensate by increased irrigation.

Comparison of model calibration results and farm data indicates that there is a high correlation between farm data and model output. The error between the model results and the Lysimeter station data is low. Among the methods used to calculate the evapotranspiration in the model, FAO Penman- Monteith method is the highest correlation and the lowest error value with the farm experiments and results. In general, the results indicated that the CERES-Wheat model has a high ability to simulate evapotranspiration and wheat yield. Regarding observed data for crop irrigation program indicates that farmers' performance in managing the amount of water needed for the crop at various stages of the growth was not optimal. Consequently, drought stress was observed for developmental and mid-growth stages. The DSSAT simulation indicated that the optimal irrigation management adjusts the time and value of irrigation water according the actual evapotranspiration and water requirement would significantly improve irrigation water use.