Rapid Determination Of Rupture Characteristic Of Large Earthquakes Based On High-rate GNSS And Accelerometer

The Wenchuan earthquake,the Qinghai Yushu earthquake,and other recent earthquakes in China have all resulted in massive loss of life and property.Earthquake prediction cannot be accomplished in a short period with current research,technology,and observation methods.In this situation,earthquake early warning,rapid postearthquake disaster assessment,and emergency response are currently the most effective ways and means of reducing the extent of damage caused by large earthquakes.The cornerstone of high timeliness applications such as earthquake early warning is high precision monitoring of near-field deformation of large earthquakes near the epicenter.In large earthquake events,inertial seismic instruments have several issues:1)Broad-band seismometers are sensitive to seismic waves but may saturate or clip with strong ground shaking near the source so that the instruments cannot record the full amplitude of seismic waves;2)Accelerometers can measure the details of strong ground motions near the source,but it is difficult to convert the acceleration data unambiguously to displacements in real-time due to tilts and rotations of the instruments.The emerging GNSS technology can directly measure the displacement and velocity waveform by relying on the satellite space datum,providing a new opportunity for high-precision seismological research.This paper aims to achieve fast inversion of magnitude and fault rupture characteristics of large earthquakes by using high-rate GNSS data in the near field of large earthquakes,to improve the ability of earthquake early warning.The methods of deformation monitoring,magnitude estimating,and fault rupture characteristics rapid inversion are mainly studied.The following are examples of specific research work and accomplishments:(1)Based on high-rate GNSS time-differenced carrier phase technology,a new data fusion method by integrating GNSS and accelerometer in a tightly coupled manner is developed.This method is less reliant on a steady base station and the precision of satellite products,and it also eliminates the requirement to estimate carrier phase ambiguity,promoting the usage of high-rate GNSS in seismology.Meanwhile,the complementary advantages of GNSS and accelerometer solve the drawbacks of GNSS,such as lower sampling rate and short-term positioning precision,and the tightly coupled method can overcome the cascade problem of loosely coupled method,which can measure accurate broad-band seismic wave signal in real-time.Experimental results show that the new tight coupled method can provide a higher precision seismic waveform than the loosely coupled method,which can not only accurately pick up the P-wave arrival time,but also accurately obtain the coseismic permanent displacement,as well as effectively solve the GNSS seismic waveform aliasing problem,and can be used as a high-precision monitoring method for near field deformation of large earthquakes.(2)Based on the velocity time series derived by high-rate GNSS time-differenced carrier phase technology,a Peak Ground Velocity(PGV)magnitude estimation method is proposed.The PGV method can accurately determine the magnitude of a large earthquake,as well as give essential parameters for earthquake and tsunami warnings.The analysis of 22 large earthquakes shows that the average absolute deviation of PGV surface wave magnitude is just 0.26,which is equal to the accuracy of the general PGD surface wave magnitude.(3)We present a new method using GNSS-derived PGVs to rapidly determine rupture characteristics for large earthquakes.A three-step strategy is adopted to sequentially estimate the fault rupture length,direction,and pattern(unilateral or bilateral).It does not require any a priori constraint on the fault plane and can avoid drift and clipping problems in seismic data based inversion methods.Through retrospective analysis of seven recent large earthquakes,the line-source rupture characteristics can be determined within 5-30 seconds.It is shown that the method has the potential to improve ground shaking predictions with applications in disaster assessment and emergency response.