A Systematic Review and Assessment of Inverse Crop Parameter Modeling Based on Synthetic Aperture Radar Data: Research advances, existing problems, and future directions

Crop canopy or plant parameters can reflect the status and trend of crop growth and are important parameters for linking crop yield and remote sensing data. Optical remote sensing data are limited by cloudy and rainy weather, and obtaining effective and continuous data is difficult. In contrast, synthetic aperture radar (SAR) data can be used for all-weather monitoring and can penetrate vegetation, facilitating in-depth analysis of crop structure and soil information. Hence, we review relevant studies on the inversion of crop canopy or plant parameters based on SAR data from 1991 to 2022. We first review the radar features and key crop parameters used in existing studies and outline the factors influencing parameter inversion accuracy. Then, we classify the inversion models into empirical, semiempirical and mechanism models and elaborate upon the principles and characteristics of each model. Finally, we discuss a series of problems in previous studies and analyze future development directions. Crop growth has artificial attributes and is more complex than natural vegetation growth. We need to integrate agronomic theory into radar mechanisms to precisely simulate microwave scattering information of crops. In modeling, we need to integrate polarization decomposition techniques and mechanism models to enhance the interpretability of SAR features. We still need to focus on the development of mechanism models to explore the "real" relationship between microwaves and crops. Additionally, we should actively investigate the integration of emerging technologies to expand the application of crop parameter inversion models, such as downscaling methods, data assimilation and transfer learning.