Abstract The retrieval of vegetation parameters benefits significantly from the data fusion of optical and microwave signals. The integration of accurate forward models in both regions can play an important role in supporting these fusion approaches. Because of the different imaging mechanisms used in optical and microwave wavelength domains, the forward models in the two domains have been generally developed separately based on the different specifications of the scene. The inconsistencies between optical and microwave models, such as confusing input/output parameter definitions, different scattering theories and discrepant model complexity, make the data fusion difficult to conduct and lead to different results in terms of accuracy and computer time. Therefore, it is of great interest to develop a unified three-dimensional (3D) model using one scattering theory, identical input and similar complexity. To our knowledge, there are very few 3D models that can accomplish this task. By extending the Radiosity Applicable to Porous IndiviDual Objects (RAPID) model for the optical region, a general radiosity model (RAPID2) was proposed in this paper for the microwave region. This is the first time radiosity theory has been applied in microwave remote sensing, which invents a new way to solve the radar multiple scattering more efficiently. RAPID2 has four new functions: projecting translucent objects, tracking specular scattering, separating polarization components and imaging radar signals. The relationship between the radar cross section (RCS) and the bi-directional reflectance factor (BRF) is bridged. The modified Stokes vector and Mueller matrix are integrated into radiosity formulas to unify the scattering process between the optical and microwave regions. RAPID2 can simulate double-bouncing and multiple scattering effects over vegetated 3D scenes containing topography. The simulated radar images can well reflect the distinct radar geometric features, including layover, foreshortening and shadows. Validation over two forest sites shows good agreement with AIRSAR backscattering data (errors

中文翻译：

---

扩展 RAPID 模型以模拟森林微波反向散射

摘要 光信号与微波信号的数据融合对植被参数的反演具有重要意义。两个区域中准确前向模型的集成可以在支持这些融合方法方面发挥重要作用。由于光学和微波波长域使用的成像机制不同，两个域的正演模型一般都是根据场景的不同规格分别开发的。光学和微波模型之间的不一致，如输入/输出参数定义混乱、散射理论不同、模型复杂度不一致等，使得数据融合难以进行，并导致精度和计算机时间方面的结果不同。所以，使用一种散射理论、相同的输入和相似的复杂性来开发统一的三维 (3D) 模型非常有趣。据我们所知，很少有 3D 模型可以完成这项任务。本文通过扩展光学区域的适用于多孔单个物体的光能传递模型（RAPID），提出了微波区域的通用光能传递模型（RAPID2）。这是光能传递理论首次应用于微波遥感，开创了一种更有效解决雷达多次散射问题的新方法。RAPID2有四个新功能：投影半透明物体、跟踪镜面散射、分离偏振分量和成像雷达信号。雷达截面 (RCS) 和双向反射系数 (BRF) 之间的关系是桥接的。修改后的斯托克斯矢量和穆勒矩阵被整合到光能传递公式中，以统一光学和微波区域之间的散射过程。RAPID2 可以在包含地形的植被 3D 场景上模拟双弹跳和多次散射效果。模拟的雷达图像可以很好地反映明显的雷达几何特征，包括重叠、透视和阴影。对两个森林站点的验证显示与 AIRSAR 反向散射数据（错误 包括停留、透视和阴影。对两个森林站点的验证显示与 AIRSAR 反向散射数据（错误 包括停留、透视和阴影。对两个森林站点的验证显示与 AIRSAR 反向散射数据（错误