Application Of High-rate GNSS And InSAR Data To Seismic Source Parameter Inversion

Seismic source parameters,such as earthquake magnitude,moment tensor,and spatiotemporal rupture process,reveal the earthquake’s physical properties,which are of great significance for understanding the mechanism of earthquake rupture,the caused damage,and the earthquake characteristics,and are important for earthquake disaster rescue and earthquake mechanism interpretation.The moment tensor and rupture process model represent the source properties and kinematics characteristics,which provide a detailed description of earthquake sources.How to quickly and stably obtain these source parameters has important scientific guiding significance for earthquake early warning and rapid response,and can deepen our understanding of the physical structure of the crust and the mechanism of earthquakes in the source area.Near-field and high-precision observations are used to constrain the source parameters with the development of modern geodesy.Near-field surface displacement time series can be solved using Global Navigation Satellite System(GNSS)technology,and the low-tozero frequency signal is reliable.Besides,GNSS data improves the efficiency of the inversion since they are distributed near the earthquake fault and can be captured in real-time.In contrast,Interferometric Synthetic Aperture Radar(InSAR)data provides intensive and high-precision zero-frequency information on near-fault surface deformation,which can be used to invert the geometric changes of complex seismic faults and deepen the understanding of the relationship between fault geometry and rupture model.The GNSS and InSAR data are the most commonly used and effective geodetic observation data for source inversion.In this study,the GNSS and InSAR data are incorporated into the source inversion and carry out a systematic study from the perspective of rapid inversion and fine inversion.The main work and contributions of this paper are as follows:(1)We performed a centroid moment tensor inversion based on high-rate GNSS waves and realized real-time inversion,the results show that source parameters based on waveforms data inversion can obtain earlier than the static data.For the 2019 Mw7.1 Ridgecrest earthquake,we analyze the centroid moment tensor inversion and its characteristics based on the teleseismic P wave,Wphase,near-field high-rate GNSS wave,and GNSS offset.The results show that this event was domain by strike-slip motion and the strike,dip,and rake angles are ～322°,～80°,and ～-176°,respectively.The fault dip angle derived from near-field data is smaller than the teleseismic data since the near-field data have a large take-off angle,resulting in a smaller inversion dip,but the results still can be acceptable.The real-time inversion results show that the moment magnitude and moment tensor solution from the GNSS wave are stable at 28 s after the earthquake’s initial time,14 s earlier than the GNSS offset inversion.It is mainly because the moving average filter is used to stable the inversion,resulting in the observation being relatively lagging than waveform data.Therefore,we conclude that the low-frequency signal in the GNSS waveform contributes to capturing the source parameters earlier.(2)We used Iterative Deconvolution and Stacking(IDS)method to realize the realtime finite-fault rupture process inversion in this study.We analyze the high-rate GNSS data’s performance,efficiency,and effect in real-time inversion by taking the 2019Mw7.1 earthquake as an example.The real-time inversion results show that in the first15-20 s after rupture initiation,the moment magnitudes in the high-rate GNSS velocity inversion are 0.1-0.3 larger than those estimated from the other three real-time inversions.The moment magnitude and shallow slip stabilized at about 30 s and 15 s from high-rate GNSS velocity waveform inversions,much earlier than the high-rate GNSS and strong-motion displacement inversions(～40 s and 20-30 s).It is attributed to the low-to-zero frequencies signal contributes to rapidly estimating the earthquake magnitude and spatiotemporal evolution.Additionally,high-rate GNSS data inversions generally need much fewer stations to generate similar real-time estimates than strongmotion data inversions.The low-to-zero frequency signals from high-rate GNSS velocity waveforms are critical for capturing the source parameters as soon as possible in earthquake early warning,which indicates a prospective potential for earthquake fast response and hazard mitigation.(3)We proposed a joint moment density tensor inversion to obtain the fault geometry and rupture spatiotemporal distribution using Interferometric Synthetic Aperture Radar(InSAR)data and seismic wave data.The 2021 Haiti Mw7.2 earthquake results show that the slip angle is ～60° around the hypocenter and dominate by the thrust motion.The slip angle is gradually reduced to ～20° in the 30 km west of the hypocenter and dominated by strike-slip motion.The fault dips in the thrust and strikeslip segment derived from our study are ～63° and ～74°,respectively,and the result is comparable to the fault geometry derived from aftershock distribution.The moment tensor inversion can obtain more fault geometry details than the traditional inversion.The kinematic rupture process also shows a unilateral rupture from east to west with an average rupture velocity of 2.6 km/s.The 2021 Haiti earthquake simultaneously ruptured at least two fault segments with varied strikes and dips,and the rupture broke through the stepover caused by the fault geometry change with a high centroid rupture velocity(5.5 km/s)and caused a cascade rupture.The rupture front experiences a directivity pulse of high ground motions with high amplitude and short duration,which may be an additional factor explaining the many landslides concentrated on the western end of the fault.