A Research On Land-Surface Snow Depth Monitoring Based On The Reflected Signals From Ground-based Multi-system And Multi-frequency GNSS Observation Network

For the high-latitude regions around the world,snow is not only a key freshwater storage,but also is an important part of global climate system.With the rapid developments of Global Navigation Satellites System Reflectometry technology,the variations of signal-to-noise ratio generated by multipath effects are used to monitor land-surface environmental parameters(including sea surface height,snow depth,soil moisture and so on)around GNSS station,whcih has become one of the research hotspots.The traditional ground techniques for snow depth monitoring have many limitations,e.g.,high-cost,low spatial-temporal resolution and poor continuity.Thus,a new GNSS-R technology based on multipath effects has been developed and applied for enhancing the warning capabilities of severe weather such as snowstorm and monitoring freshwater resource.According to the relevant researches at home and abroad,this paper analyzes the basic characteristics of GNSS-reflected signals from the correlation between multipath and SNR data,and clarifies the basic theory of GNSS-R technology.Some basic researches on the feasibility and accuracy of snow depth retrieved from GNSS-R technology have been discussed by comparing with in situ measurements.Meanwhile,a new snow depth estimation approach is proposed using the SNR combination of GNSS multi-frequency signals in multipath effects.The contents and the conclusions of this study are as follows:1.Different types of SNR signals of ground-based GNSS networks are introduced from the sevaral aspects of satellite number,carrier and signal frequency.The concept and model construction of multipath error and SNR observation are expounded,and the correlation between multipath and SNR observation is analyzed.The results show that GNSS SNR observations affected by multipath effects at low elevation angles can be used to estimate landsurface environmental characteristic parameters.2.Firstly,an overview of GNSS reflected signals and the coherent polarization characteristics are discussed.The target on filtrating the initial SNR data can be achieved based on the surrounding environment of the receiver,the signal reflection mechanism and the characteristics of the Fresnel reflection zone.For the tracks of the reflected point,on one hand,it can be used to acquire the azimuth range of monitoring the snow depth;and on the other hand,selecting the effective satellite is also available.The results show that the method for selecting SNR data can maximumlly hold the GNSS-reflected signal.3.The model on monitoring snow depth by using the GNSS-R technology based on SNR observations is introduced in this work.Then,discussing the extraction methods of the effective GNSS-reflected signals,and judging the validity of the analysis results from the LSP(LombScargle periodogram)spectrum analysis method.In addition,GNSS-R algorithm is verified by the single satellite,multiple satellites and long time series of GNSS satellites.The detection accuracy of each satellite is also analyzed.Finally,the results show that the inversed snow depth sequence has a good agreement with the in situ measurement.4.There is still a key problem that spatial resolution is poor for the existing GNSS-R methods on monitoring land-surface snow depth.Therefore,a new GNSS-R model of combining multi-system and multi-frequency SNR observation is used for land-surface snow depth detection.This model uses the clustering idea,which combines different types of SNR data to increase the number and obtain more effective GNSS-reflected signals.Eventually,the results show that the biases between the inversed results and in situ measuremnens are about 5 cm.