利用GNSS信号地表穿透特性反演土壤湿度的思考

利用全球导航卫星系统干涉信号（GNSS-interferometric reflectometry, GNSS-IR）进行土壤地表湿度反演具有无需信号源、探测区域宽广等特点，受到了国内外研究机构的广泛关注. 为解决传统GNSS穿透模型结构复杂的问题，根据几何构型给出了GNSS-IR计算天线相位中心与地表实际反射面垂直距离的方法，通过仿真不同土壤湿度下的高度测量偏差，建立了与GNSS信号理论穿透土壤深度的线性关系，以及与土壤湿度的反比例关系. 考虑到卫星高度角的变化对高度反演的影响，采用修正的测高模型，并利用Lomb-Scargle谱分析的方法计算了高度测量偏差. 为减少地表植被对反演天线相位中心距离地表实际反射面垂直距离的影响，利用归一化植被指数对高度测量偏差进行了修正，同时建立了GNSS信号高度测量偏差与土壤湿度的一阶反比例模型，最后对该模型进行了验证. 结果表明，对于长时间大尺度变化土壤湿度的场景，利用GNSS信号地表穿透特性进行土壤湿度反演有着很好的效果. 

Using Ground Penetration Characteristics of GNSS Signals to Retrieve Soil Moisture

It has been widely concerned by domestic and foreign research institutions that make the Global Navigation Satellite System Interferometric Reflectometry (GNSS-IR) used in soil surface moisture inversion because of its excellent characteristics, needing not for signal sources, suiting wide detection area, etc. In order to solve the complex problem of tradition GNSS penetration model, a method was proposed to calculate the vertical distance between the antenna phase center and the actual surface reflector by GNSS-IR according to the geometric configuration. Firstly, the height measurement deviations under different soil moisture were calculated by simulation, establishing the linear relationship of the GNSS signal with the theoretical penetration depth and the inverse proportion relationship with soil moisture. Considering the influence of the change of the satellite altitude angle on the altitude retrieval, a modified altimetry model was adopted, and the Lomb-Scargle spectral analysis method was used to calculate the altitude measurement deviation. Then, in order to reduce the influence of surface vegetation on the vertical distance between the inversion antenna phase center and the actual surface reflector, a normalized vegetation index was used to correct the height measurement deviation. And, a first-order inverse proportional model of GNSS signal height measurement deviation and soil moisture was established. Finally, verification experiments were carried out for the proposed models. The results show a better effect of the application of the GNSS signal surface penetration characteristics in the soil moisture inversion for the scene of long-term large-scale changes in soil moisture.