|Ko, Jong Han - TTU PLANT & SOIL SCIENCE|
|Maas, Stephen - TTU PLANT & SOIL SCIENCE|
Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: August 23, 2006
Publication Date: December 1, 2006
Citation: Ko, J., Maas, S.J., Mauget, S.A., Wanjura, D.F. 2006. Use of remote sensing data in modeling stressed and unstressed cotton growth. Agronomy Journal. 98(6):1600-1609. Interpretive Summary: Computer models that simulate crop growth during the growing season, or 'crop models', can be useful for predicting crop yields at harvest. But because they are based on numerical approximations of complex biological processes, they can sometimes reproduce plant growth in an unrealistic manner. Here, we field test a method that might improve these model's end-of-season yield forecasts. This method works by continuously checking the model's progress against observed crop development using remotely sensed data. If the model's progress is different from that of the observed crop at, say June 15, then the model is restarted at planting with different planting conditions. Repeating this correction method at various growth stages causes the projected end-of-season yield to more closely agree with actual yield levels.
Technical Abstract: Crop modeling and remote sensing could be combined to compensate for weaknesses of each technique. A cotton model that uses remote sensing data was tested for the possibility of using an experimental field to simulate canopy growth and to estimate lint yield using hand-held remote sensing data. Field data and hand-held remote sensing data were obtained from an experimental field at the Plant Stress and Water Conservation Laboratory at Lubbock, TX from 2002 to 2004. The model was revised so that it can be used to simulate crop growth and yield under stressed condition. We demonstrated that the model could be used for an experimental field if the model was calibrated with hand-held remote sensing data.