Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: Evapotranspiration Simulation of Winter Wheat Using Cropwat 4, Modwht 3.2, and Ceres-Wheat 3.5

Authors
item Kang, S - STUDENT - WTAMU
item Stewart, B - WTAMU
item Robinson, C - WTAMU
item Payne, W - TAES
item Evett, Steven

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: April 1, 2002
Publication Date: April 1, 2002

Technical Abstract: Crop models were proposed as a useful tool for farming management and analysis of economic risk. Performance of crop models may be affected by climates, soils and crop cultivar, etc. Three crop models, CROPWAT 4, MODWht 3.2 and CERES-Wheat 3.5, were used for model tests of winter wheat during different growing seasons at Bushland, TX, and Zhengzhou and Beijing, China. Predicted seasonal cumulative evapotranspiration (ET) by CROPWAT 4 fit the measured values better than the other two models, with small mean absolute difference (MAD) of 38 mm and root mean squared error (RMSE) of 45 mm. Predicted leaf area index (LAI) by CERES-Wheat 3.5 (MAD = 0.8, RMSE = 0.9) had a better agreement than MODWht 3.2 (MAD = 1.2, RMSE = 1.5). Both CERES-Wheat 3.5 and MODWht 3.2 had a good performance for phenological prediction (anthesis and maturity), whose MAD and RMSE were smaller than 7 d and 8.7 d, respectively. Yield simulation by CERES-Wheat 3.5 (MAD = 54 g/m**2, RMSE = 75 g/m**2) appeared somewhat better than MODWht 3.2. However, both CERES-Wheat 3.5 and MODWht 3.2 showed larger variations of yield prediction among different growing seasons and locations. Although the crop models in this study can not well simulate all crop parameters, they may be able to provide some important information for crop decision making, such as CROPWAT 4 for crop water management, MODWht 3.2 and CERES-Wheat 3.5 for phenology and yield. Additionally, the crop models require more tests and modifications for wider application.Crop models were proposed as a useful tool for farming management and analysis of economic risk. Performance of crop models may be affected by climates, soils and crop cultivar, etc. Three crop models, CROPWAT 4, MODWht 3.2 and CERES-Wheat 3.5, were used for model tests of winter wheat during different growing seasons at Bushland, TX, and Zhengzhou and Beijing, China. Predicted seasonal cumulative evapotranspiration (ET) by CROPWAT 4 fit the measured values better than the other two models, with small mean absolute difference (MAD) of 38 mm and root mean squared error (RMSE) of 45 mm. Predicted leaf area index (LAI) by CERES-Wheat 3.5 (MAD = 0.8, RMSE = 0.9) had a better agreement than MODWht 3.2 (MAD = 1.2, RMSE = 1.5). Both CERES-Wheat 3.5 and MODWht 3.2 had a good performance for phenological prediction (anthesis and maturity), whose MAD and RMSE were smaller than 7 d and 8.7 d, respectively. Yield simulation by CERES-Wheat 3.5 (MAD = 54 g/m**2, RMSE = 75 g/m**2) appeared somewhat better than MODWht 3.2. However, both CERES-Wheat 3.5 and MODWht 3.2 showed larger variations of yield prediction among different growing seasons and locations. Although the crop models in this study can not well simulate all crop parameters, they may be able to provide some important information for crop decision making, such as CROPWAT 4 for crop water management, MODWht 3.2 and CERES-Wheat 3.5 for phenology and yield. Additionally, the crop models require more tests and modifications for wider application.

Last Modified: 9/2/2014
Footer Content Back to Top of Page