不同微波遥感模态和不同数据组合的等湿度区域分布与土壤湿度的反演

论述了主动、被动和主动被动相结合的3种微波遥感模态反演土壤湿度的方法和特点。为了采用人工神经网络方法反演出土壤湿度，在随机粗糙面双谱散射模型的基础上计算了裸土壤表面的散射系数和发射率，分析了3种不同的微波遥感模态和不同数据组合的等湿度区域分布的特点，从而确定了适合于各个微波遥感模态的输入数据组合。反演结果表明，只要选择适当的人工神经网络输入数据组合，采用3种模态中的任何一种微波遥感模态反演土壤湿度都是可行的，并具有较好的反演精度，结果对于微波遥感反演土壤湿度方案的选取具有指导意义。     

从雷达后向散射系数反演土壤湿度与复介电常数

提出了一种基于小波扰动散射模型和土壤模型的从雷达后向散射系数反演土壤复介电常数与土壤湿度的方法，计算了同极化后向散射系数σｖｖ和σｈｈ之比与土壤湿度的关系曲线，给出了选择系统的工作频率和入射角的原则，分析和计算表明，这种反演方法是简单实用的有效的。     

Bi-Spectrum Scattering Model for Dielectric Randomly Rough Surface

The bistatic scattering model is often used for remote microwave sensing. The bi-spectrum model (BSM) for conducting surfaces was used to develop a scattering model for dielectric randomly rough surfaces to estimate their bistatic scattering coefficients. The model for dielectric rough surfaces differs from the BSM for a conducting surface by including Fresnell reflection and transmission from dielectric rough surfaces. The bistatic scattering coefficients were defined to satisfy the reciprocal theorem. Values calculated using the BSM for dielectric randomly rough surfaces compare well with those of the integral equation model (IEM) and with experimental data, showing that the BSM accuracy is acceptable and its range of validity is similar to that of IEM while the BSM expression is simpler than that of IEM。     

先验知识在被动微波遥感土壤湿度反演中的作用和影响

利用微波遥感的发射率数据反演裸土壤湿度，不可避免需要结合地表面和土壤层的一些先验知识，而先验知识的准确度又将对反演结果的准确度产生一定的影响。文章讨论了地表的高度起伏相关函数形式、土壤温度和土壤质地等三类先验知识，定义了几种不同的土壤湿度反演误差，从而定量地给出三类先验知识的不确定性对土壤湿度反演的影响，指出：基于BSM散射模型和人工神经网络（ANN）的土壤湿度的反演方法是可行的，向ANN输入两种极化的裸土壤表面发射率数据便可反演出裸土壤的湿度，在上述三种先验知识具有一定的不确定性时仍可保证较好土壤湿度反演准确度。     

Bi-Spectrum Scattering Model for Conducting Randomly Rough Surface

IntroductionThe scattering model for randomly rough surfacesis an important part of microwave remote sensingtheory.However,the scattering from a randomlyrough surface is a very complex problem whichdose not admit closed- formed analytic solutionsfrom the …