Research On Methodology Of Broadband Seismic Displacement Mornitoring Using A Single High-rate GNSS Receiver

Earthquakes are one of the most common natural disasters on the planet.The primary or secondary disasters caused by these earthquakes have caused huge impacts on human life and property safety.Currently,however,people's understanding of earthquakes is still in the early stage of scientific exploration,and it is impossible to avoid and correctly predict earthquakes.Against this background,if the early warning information can be issued quickly and effectively to the surrounding potentially dangerous areas after an earthquake,it can be an effective method of earthquake prevention and disaster reduction.For the earthquake early warning system,improving the accuracy and timeliness of early warning is a top priority.The existing earthquake warning systems are mainly based on the seismograph and strong seismograph,and it could perform poorly,such as underestimation of the magnitude,when large earthquakes(Mw? 7)happen.The reason is that these parameters are closely related to the surface displacements,which are vulnerable to the range limit and baseline bias from the seismometer.In addition,large earthquakes are often followed by a large number of aftershocks and early aftershocks.Therefore,a long time observation is required.On the one hand,it can effectively distinguish the surface deformation caused by the main shock,aftershocks and afterslips,thus improving the accuracy of magnitude estimation.On the other hand,it can also help us understand the process of incubation and development of the earthquake,and bring new discoveries to the geophysical research.From the perspective of early warning timeliness,the station density of the current seismic monitoring network is still not dense enough.The sparse stations will reduce the timeliness of early warning,and for large areas,the cost of building dense stations is huge.Therefore,how to quickly and accurately obtain high-frequency dynamic surface displacement and increase the density of seismic monitoring networks,has become a key issue to be solved urgently in seismic monitoring.As the high-rate GNSS can directly obtain the "absolute position" of the station relative to the global reference frame,it has been widely used in instantaneous surface displacement monitoring.It is not affected by range limiting and integration errors,and can capture lowfrequency seismic signals.Generally,GNSS network and the seismometer network are independent of each other,and the high-rate GNSS can be an effective supplement to the traditional seismologic network.Although high-rate GNSS has obvious advantages in surface deformation extraction,compared with the seismometer,there are still some shortcomings:(1)the density of the existing GNSS network is relatively sparse,leading to the low spatial resolution of seismic wave and the poor timeliness of early warning;(2)currently,the GNSS networks used for seismic monitoring are mainly equipped with GPS receivers,and the positioning accuracy and reliability is limited,making it difficult to detect the weak surface motions;(3)although the positioning accuracy of the GPS daily solution can reach mm level,its time resolution is not enough to detect deformation signals from several minutes to several hours.The high-rate dynamic GNSS has the ability to detect mid-and long-period signals,however,affected by the unmodeled errors,the 3D positioning accuracy is generally maintained at about 10 cm.What's more,unmodeled errors easily affect the identification of true ground deformation.Therefore,in-depth research on algorithms of high-rate and high-precision GPS positioning,and further exploring its potential for broadband displacement extraction,is the current hotspot in GNSS seismology and the original intention of this topic.This article aims to explore the potential of GNSS technology in the monitoring of broadband seismic surface displacements.Focus on this goal,this article mainly starts from the perspective of data processing,making an in-depth research and proposing a new method for mixed single and dual-frequency PPP to extract coseismic displacement,analyzing the benefits of multi-system GNSS to seismic monitoring in details,and proposing a solution to weaken the unmodeled errors of high-rate GNSS.The main work and contributions of the paper are as follows:(1)In order to improve the timeliness of earthquake warning and the spatial resolution of surface displacement,we developed a new approach for SF precise point positioning ambiguity resolution(PPP-AR)with mixed single-frequency and dual-frequency receivers.The ionospheric delay corrections,as a key factor for single-frequency PPP data processing,are directly derived from geometric-free phase observations with the integer ambiguity,and then interpolated for single-frequency users.Single-frequency L1 UPD parameters are also derived from the dual-frequency observation network.Using German SAPOS observation network to verify the feasibility and reliability.The results show that the accuracy of ionospheric estimation is 0.016 m,and the interpolation accuracy is 0.035 m when the station distance is 56 km.The post-process residuals of the single-frequency L1 UPD parameters are all less than 0.3 cycles,and the UPD has a high stability in the time domain,and can be provided per hour.Comparing with the daily solution,the accuracy of displacements extracted by the new method can reach1-2 cm in the horizontal direction and 2-3.5 cm in the vertical direction.The convergence time of single-frequency PPP is less than 10 min when the station distance is less than 52 km,and the convergence time gradually increases with the increase of station distance.When the station distance increases to 99 km,the ambiguity cannot be fixed,and the convergence time is greater than 20 minutes.Finally,the effectiveness of the method is verified by the real seismic event.The results show that the accuracy of the coseismic displacement extracted by the new method is basically equivalent to the fixed solution of dual-frequency PPP afterwards.(2)The rapid development of BDS and other satellite navigation systems has provided good opportunities for seismic surface displacement monitoring,rapid magnitude estimation,and fault slip inversion.Take the Mw 6.5 Jiuzhaigou earthquake as an example,we focus on the analysis of the ability of BDS and the combination of BDS + GPS + GLONASS multisystem to capture seismic signals.The results show that,in this event,the accuracies of the BDS PPP positioning are 43% and 23% higher than the GPS PPP positioning accuracy in the east and vertical directions,respectively.The velocity accuracy is comparable to GPS.Compared with the single GPS result,the accuracy of the multi-system in the east,north and vertical directions can be improved by 47%,55% and 28%.The coseismic displacements extracted by BDS and multi-system PPP are consistent with the results of the seismograph integration and the theoretical displacements,with differences in the millimeter level,which demonstrates the importance of BDS and multi-system PPP in improving the accuracy of realtime seismic wave signal extraction.(3)Phase multipath is one of the main error sources that affects the accuracy of GNSS long-term displacement.First,we introduced the mechanism of multipath effects,and several methods to reduce multipath effects.Using undifferential and uncombined PPP to extract multipath errors directly from the raw phase observations of the GPS,GLONASS,BDS,and Galileo four systems.Then,we analyze the mathematical relationship between the phase residuals between different frequencies in multi-system,and demonstrate it theoretically.The results show that the phase residuals of the BDS GEO satellites are greatly affected by the pseudorange bias.After reducing the weight of pseudorange,accurate mathematical relationships can be established between the phase residuals of all system satellites.This discovery greatly reduces the complexity of data processing.In addition,for GLONASS,this mathematical relationship can be used to recover the phase residual of R1 from R2 with an accuracy of better than 0.1 mm,indicating that the IFB has little effect on the phase residual.(4)The wavelet transform was employed to analyze the energy distribution of the multipath in different frequency bands and different spatial positions in frequency domain,thereby improving the traditional MHM model.The grid model is divided into three equal parts according to the height angle.Experiments show that the optimal grid resolution is 0.2 ° × 0.2 °× 1 °.In addition,for GPS,GLONASS,and Galileo,when the satellite elevation angle is higher than 15 °,the inter-satellite multipath has better compliance,but for BDS,it shows obvious inter-satellite differences.Although GPS L1 and Galileo E1 have the same frequency,there still exist system bias.Compared with the SF approach and the MHM model,the new M-MHM model has the most obvious improvement in positioning accuracy.The multi-GNSS PPP float solution using this method can achieve positioning accuracy of 0.75,0.55 and 2.08 cm in the east,north and vertical components,respectively.(5)The effect of CME(common mode error)on high-rate PPP is analyzed theoretically.Combined with the PCA(principal component analysis)approach,a new method for extracting CME for stations undergoing shacking is proposed.Pre-earthquake static data is used to simulate dynamic experiments,and the characteristics of CME in high-rate GNSS displacements are analyzed in both time and frequency domain.The ratio of CME and multipath error in high-rate GNSS displacements is quantified for the first time.By analyzing the seismic data,it is shown that there are long-term spurious signal in the original high-rate GNSS displacements,which will mislead the interpretation of certain geophysical phenomena.The results demonstrated that the new method can further improve the accuracy of high-rate PPP fixed solution,identify aftershocks after earthquakes,and estimate fault slip distribution and magnitude quickly and accurately.