| Soil moisture is an important parameter in the fields of agriculture,ecology and geology,and scientific and accurate monitoring of its changes is of great significance for crop growth assessment,water resource cycle and geological disaster evaluation.Global Navigation Satellite System-Interferometric Reflectometry(GNSS-IR),a new remote sensing detection technique,can realize soil moisture retrieval by separating the fluctuation term in the Signal-to-Noise Ratio(SNR).Starting from summarizing the advantages and disadvantages of existing GNSS-IR soil moisture retrieval techniques,this paper aims to improve the adaptive ability and temporal resolution of soil moisture retrieval,and conducts research on soil moisture retrieval at different time scales based on Empirical Mode Decomposition(EMD)and Cross-Correlation Satellite Selection method(CCSS).The main study components and findings are as follows:(1)Due to the limitations of using low order polynomial fitting and wavelet analysis to separate satellite reflection signals,an adaptive separation method for satellite reflection signals combining EMD and Intrinsic Mode Function(IMF)discrimination is proposed.The trend and fluctuation terms of the SNR of each satellite are obtained by this method,and compared and analyzed with the original SNR.The results show that the use of EMD can adaptively decompose the different frequency information implied in the SNR.The trend terms extracted by the IMF discriminant method are in good agreement with the overall trend of SNR.Compared with the traditional low-order polynomial or wavelet analysis,this method has stronger adaptivity and can be applied to different GNSS satellites.(2)Due to the problem that current GNSS satellite selection depends on soil moisture reference values or prior information,a CCSS method was first proposed.This method analyzes the degree of cross correlation between the phases of each satellite,and selects satellites with different observation qualities by combining them with different threshold ranges.The results show that using CCSS can not only effectively screen out available satellites,but also further screen out satellites with different accuracy levels.Moreover,this method reduces the dependence on soil moisture and achieves adaptive satellite selection.(3)In order to explore the effectiveness of GNSS-IR soil moisture retrieval at different time scales,a single-day/near-real-time soil moisture multi-satellite fusion retrieval method was constructed based on the satellite screening results of CCSS,respectively.The results show that it is difficult to accurately and timely grasp soil moisture information in the effective monitoring area near the measuring station by using only a single satellite to retrieve soil moisture.By fusing satellite phases from different directions and different time periods,the retrieval accuracy and temporal resolution of soil moisture can be effectively improved.Further comparing the satellite fusion schemes within different threshold ranges of CCSS,it is found that soil moisture retrieval results in different time scales are correlated with both the number and quality of satellites.For single-day soil moisture retrieval,when the quality of fused satellites is good,only a smaller number of satellite fusions is needed to achieve higher retrieval accuracy,and also reduce the complexity of model retrieval.For near real-time soil moisture retrieval,high dynamic monitoring of soil moisture can be achieved when the number of fused satellites is large and the quality is good. |
PDF Full Text Request