Research On Rupture And Early Warning Of Strong Earthquakes Based On High-rate GNSS And Synthetic Earthquakes

China has frequent earthquakes,bearing about 33% of the world’s strong continental earthquakes on the land area of about 7% of the world.For some small and medium earthquakes,seismic methods can quickly and accurately estimate the magnitude.For some Mw ≥ 7.0 strong earthquakes,seismic methods are easily limited by the saturation of instrument amplitude,resulting in the magnitude being underestimated.The rupture size caused by strong earthquakes often reaches tens or even hundreds of kilometers.So rapid and accurate estimation of strong earthquake magnitude and fault coseismic slip distribution is important for disaster assessment,scientific and reasonable disaster relief.In this study,we construct a set of algorithms for rapid acquisition of strong earthquake source parameters based on high-rate GNSS,and conduct retrospective tests on the constructed algorithms using the 2021 Maduo Mw 7.3 earthquake as an example to assess the applicability and effectiveness of high-rate GNSS for strong earthquake magnitude estimation and slip distribution inversion.In response to the current situation of limited strong earthquakes and relatively sparse distribution of high-rate GNSS stations in mainland China,this paper uses numerical simulations to synthesize high-rate GNSS displacement waveform data to test the constructed algorithms.The main research and findings are as follows:1.A set of algorithms for rapid acquisition of strong earthquake source parameters based on high-rate GNSS data,from processing of high-rate GNSS data,estimating the magnitude of strong earthquake,and inverting the slip distribution.The GNSS horizontal positioning accuracy reaches the mm level,and the vertical positioning accuracy is better than 2 cm.The estimation of peak ground displacement(PGD)and peak ground velocity(PGV)based on velocity and displacement waveforms has been achieved,and the displacement waveforms can be used to invert the spatial and temporal processes of coseismic slip on fault surfaces.The specific results are as follows:(1)The horizontal positioning accuracy obtained by the relative positioning method can reach mm level and the vertical positioning accuracy is better than 2 cm when processing the high-rate GNSS data.The accuracy of horizontal velocities obtained by the variometric method can reach mm/s,and the accuracy of integration into horizontal displacements can reach mm level,which can provide the basis for rapid acquisition of coseismic deformation.(2)Based on the displacement and velocity waveforms of high-rate GNSS,we construct an algorithm for PGD and PGV magnitudes.Based on the displacement waveforms of high-rate GNSS,the steepest descent method can be used to invert the magnitude and fault slip distribution results at each epoch given the threshold value of the seismic event-triggered deformation magnitude.2.The constructed algorithm for rapid acquisition of seismic source parameters based on high-rate GNSS was retrospective tested by the high-rate GNSS data from the 2021 Maduo Mw 7.3 earthquake.Results show that the displacement accuracy of the high-rate GNSS data can reach the mm level,the estimated PGD and PGV magnitudes are consistent with those obtained by seismological methods,and the slip distribution results are consistent with the post-processing results in terms of rupture size and slip.The specific results are as follows:(1)The accuracy of the displacement and velocity results obtained by using the relative positioning method and the variometric method can reach the mm level,and the fluctuation amplitude of the displacement and velocity waveforms gradually decreases with increasing source distance.PGD and PGV magnitudes of this earthquake estimated using displacement and velocity waveforms are Mw 7.25 and Mw 7.31,respectively,which are consistent with the seismological method(Mw 7.3).(2)Results of the fault slip distribution obtained by retrospective inversion of the horizontal displacement waveforms indicate that we can acquire an initial inversion magnitude(Mw 6.65)34 s after the earthquake,and the magnitude gradually increased with time and converged to Mw 7.40 61 s after the earthquake.There is a large difference between the maximum slip obtained from our inversion and the maximum slip obtained from the inversion using near-field InSAR or dense GNSS data,but our results are consistent with the post-processed results in terms of magnitude and rupture size.Results indicated that the GNSS displacement waveforms in the near field can effectively characterize the release process of the seismic moment.(3)We use near-field dense high-rate GNSS displacement waveforms to invert coseismic rupture process,obtain the ground motion parameters by using forward simulation method and further convert to the seismic intensity.Although there are differences between the intensity results obtained from simulation and field survey,they have first-order characteristics in the extreme seismic zone.Results indicate that the rapid inversion of coseismic-related parameters based on high-rate GNSS can provide a reference basis for post-earthquake rapid rescue and disaster investigation.3.In view of the limited number of strong earthquakes and few high-rate GNSS stations,we synthesize 2100 Mw 6.0-8.0 earthquakes in Sichuan-Yunnan regions based on the numerical simulation method of coseismic rupture with the interseismic locking model as a priori constraint.We use forward simulations to synthesize theoretical displacement waveforms at existing GNSS stations and test the “rapid acquisition of strong earthquake source parameters based on high-rate GNSS data”.The statistical results show that the existing GNSS continuous stations around the Anninghe fault were able to determine the initial inversion magnitude(Mw 6.2)about 8 s after the event,indicating a greater advantage of high-rate GNSS for estimating the magnitude and slip distribution of strong earthquakes.The specific results are as follows:(1)Synthetic earthquakes based on the interseismic locking model found some deviations between the PGD/PGV magnitudes and the actual input magnitudes,but the average results from multiple stations can recover the actual input magnitudes better.(2)For the Mw 6.5-7.5 earthquakes on the Anninghe fault,where GNSS stations are dense,the initial inversion magnitude(Mw 6.2)can be determined about 8 s after the earthquake using the existing GNSS continuous stations,and accurate magnitude estimates and slip distribution results can be obtained within a short time or before the rupture is completed.(3)GNSS stations to be built at the “Earthquake Science Experimental Site” can provide a more effective picture of the magnitude estimation,slip distribution and even the dominant direction of rupture,indicating that high-rate GNSS has a greater advantage in estimating the magnitude and slip distribution of strong earthquakes.4.For historical strong earthquakes,there are many gaps in the understanding of rupture patterns and slip characteristics due to limited observational data.While carrying out numerical simulations of synthetic earthquakes,we explore the coseismic slip characteristics of the 1850 Xichang M7.5 earthquake by combining the present-day deformation,numerical simulations,geomorphic dislocations and investigated intensities.Results show that this earthquake may have ruptured only the northern section of the Zemuhe fault,with the main slip located above 15 km depth.The specific results are as follows:(1)We select 29 slip distributions among the stochastic rupture scenarios synthesized in the northern section of the Zemuhe fault,and the offsets generated can be consistent with the geomorphic offsets results obtained from fieldwork measurements,and the simulated peak ground acceleration(PGA)distribution is consistent with the surveyed intensity results.(2)The slip distribution of these 29 events was averaged to obtain the slip distribution of the fault plane,which roughly depicts the first-order characteristics of the slip distribution of the 1850 Xichang M 7.5 earthquake,i.e.,the main slip zone is located at 0-15 km depth and the rupture may extend to the surface.(3)The 1850 Xichang M 7.5 earthquake may have ruptured only the northern of the Zemuhe fault,but not the Anninghe fault.The Xichang transition zone located between both faults may have prevented rupture propagation to the Anninghe fault.The significant difference in the strong ground motions caused by the single-segment rupture and multi-segment rupture patterns of the 1850 Xichang M 7.5 earthquake is indicative of the importance of different rupture patterns in the assessment of strong earthquake hazard.In this study,we construct a rapid acquisition algorithm for strong earthquake source parameters by using high-rate GNSS data,and test this algorithm using high-rate GNSS data from the 2021 Maduo Mw 7.3 earthquake and synthetic earthquake waveforms,which is an important reference for future disaster assessment and post-earthquake rescue during strong earthquakes.In addition,we explore the coseismic slip characteristics of the 1850 Xichang M 7.5 earthquake based on interseismic locking characteristics,fusing geomorphic data with numerical simulations,which will provide new ideas for exploring other historical earthquake slip characteristics in the Tibet Plateau,and provide some new insights for regional disaster planning.