Research On Monitoring Rain Rate Using Polarimetric Phase Shift Of GNSS Signals

With the rapid development of the Global Navigation Satellite System(GNSS),the GNSS remote sensing technology of atmospheric and oceanic environment,which aims to fully tap the value of GNSS big data,has attracted extensive attention,and some technologies have been implemented in business applications.This thesis focused on the critical technique and scientific topic on detecting rain rate based on polarimetric phase shift of GNSS signals.Mainly,theoretical research on forward model and inversion algorithm,experiment verification research on development of experimental equipment,experimental design,data processing algorithm and comparative analysis of the results,simulation research on retrieval of regional rain cluster information were deeply analyzed.This study would make important contribution to the independent development of China's GNSS signal telemetry rainfall system.In theoretical research,based on the physical process of GNSS signals propagating through rain-filled medium,a relation model between polarimetric phase shift and rain rate,namely the forward model was established.Then,the relation model was simulated to analyze its feasibility of estimating rain rate using this model.And,the effects of raindrop size distribution,scattering algorithm of nonspherical particles,shape of raindrops,raindrop canting angle,rain path length,temperature,frequency and elevation angle of GPS satellite were systematically investigated.The results show that this relation model has the possibility to estimate rain rate and different factors have different effects.For the inversion problem,a Bayesian method based on the simulated annealing algorithm was proposed and verified by simulated data.In addition,the effects of other factors in the slant path such as ionosphere,ice crystals and melting particles were also researched,drawing a conclusion that ionosphere effects are not obvious,the effect of ice crystals can be negligible,while that of melting particles should be considered.In experiment verification research,firstly,on the basis of the above theoretical research,a conical horn antenna used for receiving GNSS dual-polarized signals was developed,and a GNSS dual-polarized rain detection system was constructed.Then,taking the GPS signal as an example,twice ground-based experiments were conducted on June-September 2015 and July-November 2016 to monitor rain rate using polarimetric phase shift.Secondly,considering that the raw data of experiment was dual-polarized GPS carrier phase,the data processing algorithm to solve many problems such as the quality control,lost lock problem,cycle slip problem,ionosphere,neutral atmosphere,hardware effects,uncertainty of initial value,multipath effect were researched and proposed.Also,data processing method for comparative analysis of multi-source data was studied.Based on the case analysis and statistical analysis of data from 2 satellites in 2015 and 13 satellites in 2016,the following conclusions can be drawn: the polarimetric phase shift can be obtained and it can be determined that the polarimetric phase shift is caused by rainfall;the polarimetric phase shift has a good correlation with the rain gauge data from some meteorological stations and the path-averaged rain rate retrieved from radar,but the correlation between polarimetric phase shift and the path-averaged rain rate from radar is significantly higher than that with rain gauge data;compared one hour cumulative polarimetric phase shift of 29 cases with one hour cumulative rainfall from 4 nearby meteorological stations respectively,the optimal correlation coefficients are all more than 0.96,even some reaching 0.99.These results fully verify the effectiveness of the data processing algorithm and the credibility of the polarimetric phase shift.In regional simulation research,in order to solve the ambiguity between rain path length and rain rate and get the regional rainfall information,a tomographic model to obtain two-dimensional structure of rain clusters retrieved by polarization phase shift of one-dimensional GNSS occultation signals was established based on Computer Tomography(CT)technique.An inversion scheme using two regularization algorithms(Truncated Singular Value Decomposition(TSVD)method,Tikhonov method)and two regularization parameter selection methods(L curve criterion and Generalized Cross Validation criterion)was proposed.Finally,numerical simulations were performed to investigate the feasibility and effectiveness of this scheme under overdetermined and underdetermined conditions respectively.This research makes a helpful exploration for the subsequent data processing and applications of space-based data.