高频GPS和强震仪组合观测网络在地震预警和快速响应系统中的应用

With the rapid development of geodetic techniques, high-rate GPS(GNSS) has became a powerful complementary measurement to traditional seismic networks in monitoring dynamic ground deformations in real time. We investigate the applications of joint high-rate GPS and strong-motion networks in earthquake early warning and rapid response from four progressive perspectives, including the determination of high accuracy real-time ground motions, the picking-up of seismic phases and the location of hypocenter, the earthquake early warning based on joint measurements, the rapid determinations of earthquake source parameters in rapid response. After discussing different methods of integrating observations of collocated high-rate GPS and accelerometer, the real-time dynamic coseismic deformations with the accuracy of millimeter are achieved. The optimal method is investigated to pick up seismic phases from real-time waveforms determined by high-rate GPS-only network, accelerometer-only network and joint networks, respectively. The earthquake early warning parameter suitable for moderate and great earthquake based on the joint measurements is proposed. Considering the uncertainties of fault geometry and other nonlinear parameters, the real-time algorithms to determine earthquake parameters for point, line and dislocation sources are developed. The rapid inversion method for earthquake rupture process is also explored in the mixed time/frequency domains. At last, an earthquake early warning and rapid response application system based on the joint high-rate GPS and strong-motion networks is developed. The project shows great scientific significance to the progress of earth science and practical significance to disaster prevention and mitigation.

随着空间大地测量观测技术的发展,高频 GPS(GNSS)已经逐渐成为动态地震地壳形变监测的主要手段，也成为了传统地震观测台网不可或缺的补充手段。本课题从实时动态地表形变监测、地震震相拾取、地震预警参数确定和震源参数的快速确定四个方面系统地研究高频GPS和强震组合观测网络在地震预警和快速响应的应用。研究并址高频GPS和强震仪观测数据数据融合方法,获取毫米级定位精度的动态同震形变场;研究从高频 GPS 观测网络、强震台网及其组合观测网络中有效实时拾取不同震级地震震相的方法;基于高频 GPS 和强震仪组合网络的观测数据研究普适的地震预警参数;研究顾及非线性参数的点、线、面地震源动态位错最佳实时反演算法，采用混合时域、频域分析方法对震源破裂过程进行快速反演；开发一套基于高频GPS和强震观测网络组合定位的地震预警和快速响应系统软件。本项目研究成果对发展地球科学理论和防灾减灾具有重要的科学与现实意义。

随着空间大地测量观测技术的发展，高频GPS逐渐成为动态地震地壳形变监测的主要手段。本项目针对高频GPS和强震仪组合观测在地震快速预警中的关键理论方法和关键技术问题，分四个方面开展了深入研究，包括（1）高采样GPS和强震观测网络实时定位数据处理与方法；（2）地震波震相识别方法；（3）地震预警参数；（4）快速响应确定震源参数方法。具体研究了（1）多尺度多路径主成分分析降噪方法、基于完备总体经验模态分解的多尺度多方向去噪方法、CEEMD-MPCA阈值去噪方法、高频GNSS与强震仪的实时松/紧组合方法、顾及有色噪声的高频GPS和强震仪数据组合定位、基于抗差Sage滤波的高频GPS与强震仪数据融合处理；（2）基于长短时窗法和AIC方法的地震波震相识别；（3）利用高频GNSS用于提高预警震级的实效性和可靠性的地震预警震级估计方法；（4）利用高频GPS波形反演2016年意大利地震序列和2019年Mw 7.1加州Ridgecrest地震震源参数、基于方差分量估计2015年尼泊尔Mw 7.8地震同震滑动分布。研究结果表明：（1）高频GPS时间序列去噪新方法，能有效削弱频率范围在0.00025-0.5 Hz的高频GPS信号误差，均方根误差减少约60%；（2）顾及有色噪声的融合模型可以使高频GPS和强震仪数据松组合更为稳定可靠，能更有效地拾取P波到时且明显改善基线漂移问题；（3）ACI方法是一种准确度最高的震相识别方法；（4）地面峰值位移（PGD）可以快速可靠地确定地震预警震级，而Variometric方法能获得与精密单点定位相同的震级预报水平；（5）通过2～3个高频GPS台站数据可得到较为可靠的震源机制解，显示出利用高频GPS进行地震预警的巨大潜力；（6）在小江断裂带布设3台地震大地测量仪（SG160-09），构建地震预警平台。本项目的研究成果为后续地震科学研究和地震预警监测提供了重要的理论和技术支撑。