Abstract:

As an important factor for evaluating wheat quality, grain protein content (GPC) is crucial for guiding agricultural production and enhancing the market value of wheat. To advance the development of GPC remote sensing monitoring techniques, this paper systematically summarized the latest research, with a particular focus on analyzing the strengths, weaknesses, and challenges of diverse GPC remote sensing monitoring models. Results showed that remote sensing data from various platforms−including ground, unmanned aerial vehicles (UAVs), and satellites−each exhibit distinct advantages in monitoring GPC in wheat. However, as data scalability increased, the accuracy of GPC monitoring tends to decrease slightly. In terms of model construction, the development of wheat GPC monitoring models from empirical models to semi−empirical models or coupled remote sensing and crop growth models had increased agronomic parameters and ecological factors, which effectively improved both accuracy and spatio-temporal scalability. It was shown that the semi-empirical models were the preferred option for monitoring GPC. After adding meteorological factors into the Beijing wheat GPC model that integrated spectral information and agronomic parameters, the model's R² increased by 0.242. Currently, there were still many challenges in terms of model accuracy and regionally applications such as the reliability of GPC data, the complexity of the vertical distribution of nitrogen in wheat, and the limitations of regional expansion of the models. To address these issues, this paper proposed to evaluate ground-based GPC, fusing effective data, mine spectral information and explore multi-scale transformation methods in the future. In addition, a multi-scale GPC monitoring model based on collaborative observations from ground stations, UAVs and satellites can be constructed to achieve efficient, accurate and comprehensive monitoring of wheat quality.

Key words: Wheat, Grain protein content, Remote sensing monitoring