Key points are not available for this paper at this time.
Crop yield prediction plays an important role in food security. Crop growth models can be used for this purpose, however, they require empirical parametrization. In this study, we show that remote sensing observations can be used to parameterize a crop growth model, providing crop growth monitoring and yield predictions, while mitigating the need for extensive field measurements and accounting for variability in crop varieties. For this purpose, we coupled the SCOPE radiative-transfer model with the state-rate crop model WOFOST. In this paper, we explain how the two models are coupled, assess the accuracy and present a global sensitivity analysis of the coupled model. To assess the accuracy of the coupled model in predicting growth and yield, experimental data of five potato varieties from two fields, in two different locations, in the Netherlands were used. Results show that simulated LAI had a good correlation with measured LAI for four out of the five varieties (R2 > 0.6), while RMSE was in the range 1.09–1.34 m2 m-2. Simulated leaf dry matter content showed similar results, with high accuracy for all varieties (R2 > 0.6) and RMSE between 310 and 440 kg ha-1. Finally, yield simulation showed promising results with an average difference of 1800 kg ha-1 between simulated and measured values. Based on the findings of this study, it is possible to conclude that a coupled WOFOST-SCOPE model has the potential to improve the application of remote sensing data for crop growth monitoring and yield prediction.
Ntakos et al. (Sat,) studied this question.