Research On Retrieving Water Level,Snow Depth And Soil Moisture Using GNSS-reflected Signal

The changes of snow depth,soil moisture and other surface parameters have huge impact on the climate,the environment and the survival of mankind.At present,the monitoring of snow depth and soil moisture mainly relies on in situ measurement and satellite radar remote sensing,which have rspective deficiencies.In situ measurement requires much manpower and material resources,and single station can only observe a rather small area(e.g.a few square meters).The satellite radar remote sensing methods generally are higher cost and have lower spatial resolution and longer visit time,which cannot meet the requirements of real-time continuous monitoring.As the global navigation satellite system(GNSS)increasingly becomes mature,it has gradually attracted wide attention in the field of remote sensing due to its unique advantages:(1)the use of passive signals without transmitter greatly reduces the complexity,size and cost of the remote sensing system;(2)it can be used in a wide range of applications,such as inversion of sea surface wind field,seawater flow field,salinity,sea ice,sea surface elevation,surface soil moisture and snow depth;(3)a large number of signal sources can achieve all-weather uninterrupted global coverage,which is conducive to the realization of data acquisition and target inversion with low-cost,large-scale,high space-time resolution and can also fill the mesoscale resolution bank to existing means of observation.This thesis aims to study the key techniques and methods of retrieving three surface parameters(reservoir water level,snow depth and soil moisture)systematically based on GNSS-R technology.In this paper,we proposed three new surface parameters inversion methods by using the theoretical derivation,modeling analysis and experimental comparison,icluding reservoir water level mornitoring using the GNSS deformation monitoring system observation data,snow depth inversion using the trilple phase combination observation,and the inversion of surface soil moisture based on the GEO-R and GEO-IR technology.The main contents and contributions of the paper are summarized as follows:1)Study the structure and characteristics of GNSS signals.Systematically sort out the constellation composition and the basic composition of the broadcast signals of each GNSS system,respectively.Using the electromagnetic wave scattering theory,including polarization and reflection characteristics,the basic model of received GNSS direct signal and reflected signal are derived.Based on the signal model and feature,the capture and processing of reflected signal are described in detail.Finally,the plane and the spherical geometry and the reflection principle of GNSS-R/GNSS-IR as well as the corresponding observation and application are systematically introduced.2)Based on the theory of GNSS-R,a method is proposed for monitoring the reservoir water level using the observation data provided by the existing GNSS deformation monitoring system of reservoir dam.The basic principles and methods of GNSS-R and GNSS-IR altimeter are analyzed in detail,and then the water level of two reservoirs of Shanxi Xilongchi and Shenzhen Kengxi are used to analyze and verify this method.The results of Xilongchi show that the accuracy of signal-to-noise ratio method is slightly higher than that of the dual-frequency phase combination method.However,because of the elevation angle limitation,the time of obtaining the real water level and the acquisition time of the GPS data are inconsistent,resulting in lower accuracy of inversion,with the mean square errors being 0.312m and 0.326m,respectively.Under the condition that the real water level observation time coincides with the GNSS data acquisition time,the root mean square errors of the inversion of the water level with the three GPS satellites(G10,G22 and G31)are 3.88cm,3.38cm and 3.96cm,respectively,achieving cm level accuracy.3)The snow depth estimation method using triple frequency carrier phase combination observations is.proposed.The signal-to-noise ratio(SNR)and geometry-free phase combination(L4)are the two basic methods for the inversion of snow depth.The SNR method needs to remove the trend term and influenced by the quality of the observed data.Although the L4 method is not affected by the geometric distance,it can not reduce the influence of the ionospheric parameters.Aiming at dealing with these defects and shortcomings,the method of snow depth eatimation using triple frequency phase observations is proposed,which can eliminate the impact of both geometric distance and ionosphere.Two experiments were carried out to verify this method,and the inversion results are compared with the SNR method and L4 method.The results show that the inversion precision of triple frequency combination method can be as high as subcentimeter,superior to SNR method and L4 method.4)The method of soil moisture inversion by GEO-IR is proposed.The existing MEO-IR inversion method of soil moisture has the problem of limited observation time,while the observation time of the earth synchronous orbit(GEO)satellite is uninterrupted.In order to make full use of the characteristics of GEO satellite,two methods are proposed for the retrieval of soil moisture based on SNR and phase combination,respectively.The GEO-IR method can make full use of the observation data of the continuous operation reference stations all over the world.It can improve the soil moisture inversion resolution,and also achieve all-weather soil moisture monitoring.Then,the method is verified by simulation and experimental data.The results showed that the soil moisture inversion accuracy of the GEO-IR method is comparable to that of MEO-IR,reaching 0.003cm3 · cm-3.5)The method of soil moisture inversion by GEO-R is presented.The method uses the reflected signal time delay Map(DM)and the ratio of the reflected signal SNR and the direct signal SNR.Compared with MEO-R based time delay Doppler Delay Map(DDM)method,the GEO-R method can improve the extraction accuracy of maximum power and improve the accuracy of soil moisture retrieval.At the same time,the method does not involve the Doppler frequency,simplifying the data processing and reducing the transmission consumption.Based on the GEO-R theory,the DM maximum power and SNR ratio are modeled with the linear and second-order parabolic inversion models of soil moisture,and the results are verified by simulation and experimental data.As for the DM maximum power method,the accuracy of the linear model and that of the second-order model are respectively 0.038cm3 · cm-3 and 0.01cm3 · cm-3.The linear model of the SNR method has accuracy 0.0231cm3 ·cm-3,which is significantly better than that of the DM method,while the second-order model of the SNR method achieves accyracy of 0.0201cm3 ·cm-3,which is much worse than the counterpart of the DM method.