Soil Moisture Retrieval Using BDS Signal-to-noise Ratio Observations

Soil moisture is a significant physical quantity that can characterize the soil conditions.It is one of the important factors that determine drought,river runoff,and vegetation cover,and it plays a very important role in climate,hydrology,agriculture and disaster warning.Therefore,it is of great significance to carry out research on soil moisture retrieval.Traditional measurement methods and remote sensing cannot meet the requirements with high spatial and temporal resolution.Soil moisture retrieval based on Global Navigation Satellite System – Interferometry and Reflectometry(GNSS-IR)can be applied on the existing GNSS networks to solve these problems and has the great potential to complement existing soil moisture monitoring networks.With the third generation of Bei Dou satellite navigation System(BDS-3)officially completed and fully operational,soil moisture monitoring using GNSS-IR based on the BDS has great application value under the background of satellite navigation,signal science,soil science,and remote sensing.In this paper,the soil moisture retrieval research based on BDS is carried out from the analysis of reflected signal,the signal processing,and the construction of retrieval model.This paper discusses the optimization strategies of the key parameters and the models based on the traditional MEO satellites,constructs three different orbit satellite retrieval models of BDS GEO,IGSO,and MEO,and studies the refined soil moisture retrieval method that takes into account the contribution differences of the reflected signals.The main content and achievements of this paper are as follows:(1)The basic theories and methods of soil moisture retrieval using GNSS-IR are studied.Based on the analysis and comparison of the satellite constellations and orbit types of the four major GNSS satellite navigation systems,the vector relationship of the signal-to-noise ratio,the geometric relationship and polarization characteristics of the reflected signals,and the related theories and methods of the Fresnel zone are analyzed in detail.(2)The optimization strategies of the key parameters and the models based on the traditional MEO satellites are compared and analyzed.From the experimental point of view,the influences of different choices on the accuracy of the results are analyzed based on three parameters(effective antenna height,satellite elevation range,and signal frequency)and two traditional algorithm models(satellite track model and daily average model).The results show that selecting data within an elevation angle range of [5°,30°] and an arc length greater than 15° can achieve better results,the selection of the priori reflector height does not have an evident impact on the results,the L2 signal is better than the L1 signal for soil moisture retrieval,and the satellite track model and the daily average model are roughly equivalent in accuracy.(3)Soil moisture retrieval models for BDS GEO,IGSO,and MEO satellites are constructed.Based on the analysis of the orbital characteristics and signal-to-noise ratio observations characteristics of the three different orbiting satellites,the retrieval models are constructed.The determination coefficient between the results calculated by the second-order model of the BDS IGSO satellites and the soil moisture reaches 0.6417,and the determination coefficient calculated by that of BDS MEO satellites reaches0.9858.For the BDS GEO satellite,the determination coefficient between the results calculated by the second-order model and the soil moisture reaches 0.8286.The comparison of the results between the three retrieval models and the traditional GPS MEO satellite second-order model are as follows: the BDS MEO satellite model is the best,the GPS satellite model and the BDS GEO satellite model are the next,and the BDS IGSO satellite model is the worst.(4)The refined soil moisture retrieval method that takes into account the contribution differences of the reflected signals are studied.A regional refinement method based sector and grid is proposed,which can reflect the change of soil moisture within different distances from the station.The results show that the determination coefficient between the results calculated by the model and the soil moisture reaches 0.7466,and the average RMSE and MAE are 0.0260 and 0.0219.The method can effectively reflect the change of soil moisture within different distances from the station,and it can improve the practical application scenario of the method.