Soil Moisture Inversion Based On Beidou GEO Satellite Signal-to-noise Ratio

As an important part of the land water cycle,soil moisture is a great significance to agriculture,climate,disaster prevention and control,etc.Conventional methods of measuring soil moisture,such as humidity measurement or weighing,can measure soil moisture with higher accuracy,but in the process of measurement,not only destroys the structure of the soil itself,but also wastes a lot of manpower and material resources.GNSS-R technology is an emerging method for measuring soil moisture.Because it has a large number of free L-bands,it has wide coverage and strong penetration,and it does not require a special launch source to have a unique advantage.In the application of GNSS-R technology in soil moisture inversion,GPS is relatively mature for the inverting of double antennas and single antennas,but it is impossible to invert and monitor soil moisture in fixed areas.In this thesis,GEO satellite in geosynchronous orbit is used for experiments.Because its position is stationary relative to the position of the receiver and the satellite elevation angle and azimuth angle are fixed,the Beidou GEO satellite can be used to inversion the soil humidity in the fixed area.During the all-weather observation period,when soil moisture mutates due to rainfall or other factors,the observation data is inaccurate and affects the final observation results.In view of the defects of previous experiments,this thesis proposes a method combining characteristic time periods with random time periods.The experimental data are collected before,after rain and when soil moisture is relatively stable,and 60 groups of observations are obtained before and after.The field determination of receiver position and ground reflection point directly affects the acquisition results of soil moisture truth value.The receiver position is calculated by GAMIT/GLOBK baseline calculation network adjustment,and the accuracy is millimetre.Using the S.C.WU algorithm to iterate the coordinates of the mirror reflection point,and conducting field experiments to obtain the seven parameters of the WGS84 coordinates of the test area and the Beijing 54 coordinate conversion,the Beijing 54 coordinates of the mirror reflection point are obtained through the conversion of the coordinates,and the conversion accuracy is in centimeters.The field location of the mirror reflection point was finally determined.The Fresnel principle,Fresnel space and ground reflection model are studied.According to its theory,the effective inversion area based on the soil moisture inversion of the Beidou C01 satellite is obtained.When the receiver height is 1.46m,The fixed area with an area of 22.4m² can be inversion,that is,when combined with five GEO satellites at the same time,soil moisture inversion can be performed on the fixed area range of about 100m².In the experiment,sinusoidal fitting and smooth-spline curve fitting are used to model the multi-path components respectively.For the multi-path component of sinusoidal fitting,the amplitude feature is extracted,and the multi-path component of smooth curve fitting is extracted.The overall average value,the maximum value average value,and the minimum value average value are extracted.There are a total of four eigenvalues.Through the analysis,the overall mean characteristics can only reflect the concentration trend of multi-path components on the one hand,and there is no obvious relationship with the change of soil humidity.The order of magnitude basically tends to 10^-2,reflecting the periodicity of multi-path components.Positive and negative symmetry.The other three characteristic values have obvious linear correlation with the true soil moisture value.Linear fitting of the other three characteristic components with the real soil moisture values was carried out,and the determining coefficients of the three fitting were all close to 1.The determination coefficient of amplitude component is 0.9363,which is closest to 1.Therefore,the linear fitting of amplitude component and soil moisture is the most powerful explanation for the sample data,followed by the maximum value component.