Predicting certain crop phenological stages is important for scheduling agricultural practices and predicting crop responses to climate change. In this study, we developed three different wheat phenological models, a polynomial model and two sigmoid and exponential mixed SEM models developed by different parameter determination methods the Nelder-Mead and augmented Lagrange multiplier methods , and determined which of these models is the most effective for predicting the flowering date in wheat. Five winter wheat cultivars were cropped in western Japan for four years; we split the cultivation data for model calibration and validation. The SEM models showed higher precision in root mean square error RMSE; 3-5 days than the polynomial model when using the validation data. The models developed using the Nelder-Mead and augmented Lagrange multiplier methods showed similar RMSE values Mean SD: 4.24 0.59 and 4.16 0.36, respectively . On the other hand, in the context of validity, the model developed using the Nelder-Mead method showed an unnatural development response to changes in environmental variables; thus, we found that the model developed using the augmented Lagrange multiplier method would be more realistic and effective to express the response of wheat growth to environmental factors. The results of our study shed new light on the optimization methods used in crop development models and on the advantages of using the augmented Lagrange multiplier method for determining the parameters of a non-linear crop development model.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.