Feasibility of yield estimation based on leaf area dynamics measurements in rice paddy fields of farmers

Rice yield is an important factor for farmers when reviewing crop management planning and cultivation operations. In previous studies, a non-destructive, labor-saving method has been proposed to determine yield using spectral information from remote sensing. This method assumes that molecule number (e.g., chlorophyll content in leaves) and biophysical information (e.g., leaf area) cause spectral variations. However, spectral information cannot reflect the influence of physiological factors because of the limitations of this method. In this study, we investigated solar radiation use efficiency (RUE), which reportedly reflects physiological changes in crops. Time-series data of leaf area index were collected over five years in rice paddy fields of farmers to calculate and investigate the variation in RUE. The results showed that RUE varied among the survey points but did not differ significantly by year. The mean absolute error of yield estimation ranged from 68.2 g m−2 to 102.2 g m−2 depending on the yield estimation method and from 36.8 g m−2 to 454.1 g m−2 depending on the RUE. Among the estimation methods, the highest accuracy was achieved using leaf area dynamics. These results provide insights into both the effectiveness of leaf area dynamics in yield estimation and the potential errors when using remote sensing. •Solar radiation use efficiency differs by location, planting method, and year.•Error in yield estimation using leaf area differs by solar radiation use efficiency.•Leaf area dynamics is effective in improving the accuracy of rice yield estimation.