GNSS-IR双频数据融合的土壤湿度反演方法

目前全球导航卫星系统反射信号干涉测量（GNSS-IR）土壤湿度反演研究仅针对单一频点展开，提出用熵值法将2个频点数据进行融合以改进土壤湿度反演精度。首先，利用频谱分析法分别解析出各频点的信噪比（SNR）序列的振荡频率，计算出对应的等效天线高度，并利用最小二乘法求解各频点信噪比序列相位；然后，通过熵值法进行2个频点的相位观测量融合；最后，利用融合结果与实测土壤湿度建立经验模型，实现土壤湿度反演。利用地基观测实验获得的全球定位系统（GPS）L1和L2信噪比数据对该方法进行了验证，结果表明:L1和L2双频融合反演结果平均标准差为0.6%，比L1单频反演结果提高64.73%，比L2单频反演结果提高32.12%；均方根误差为0.37%，比L1频点降低72.8%，比L2频点降低73.4%。 

Soil moisture inversion method based on GNSS-IR dual frequency data fusion

At present, the study of soil moisture inversion in the field of global navigation satellite signal-interferometer and reflectometry (GNSS-IR) is only for single frequency deployment. In the paper, we propose a method that uses the entropy method to fuse two frequency to improve the accuracy of soil moisture inversion. First, the spectrum analysis method is used to analyze the oscillation frequency of the signal-to-noise ratio (SNR) sequence of each frequency point, and calculate the corresponding equivalent antenna height. The different frequency phase of SNR sequence can be solved by least square method. Then, the phase observation of two frequencies is fused by the entropy method. Finally, an empirical model was established by using the fusion results and the measured soil moisture to achieve soil moisture inversion. The method was verified by global positioning system (GPS) SNR ratio data obtained in frequency L1 and L2 by ground-based observation experiments. The results show that the average standard deviation of the L1 and L2 inversion results after dual frequency fusion is 0.6%, which is 64.73% higher than the L1 frequency inversion results and 32.12% higher than the L2 frequency inversion results. And the RMSE is 0.37%, 72.8% lower than L1 frequency and 73.4% lower than L2 frequency.