Abstract: Earth-based radar (EBR) serves as a crucial supplementary tool for lunar exploration. However, the complex variation of incidence angles during lunar observation poses challenges for interpreting radar remote sensing data. The existing empirical lunar echo model, known as the two-component model, exhibits limited accuracy and wavelength adaptability. To address these limitations, we propose improvements through function modification and process optimization of the model. Validation using 4.2 cm and 60 cm wavelength datasets demonstrates that the modified model better captures incidence angle-dependent characteristics of backscatter coefficients and polarization ratios, showing closer alignment with experimental measurements and theoretical predictions. These results confirm that the accuracy and wavelength compatibility of the model is enhanced. This study provides significant implications for data interpretation and feature extraction in EBR-based lunar exploration.