The Case for Collocation of GNSS and Seismic Networks for Earthquake and Tsunami Early Warning

Session: Advances in Real-Time GNSS Data Analysis and Network Operations for Hazards Monitoring
Type: Oral
Date: 4/30/2020
Time: 01:30 PM
Room: 230 + 235
Description: 

The utility of GNSS networks for earthquake early warning has been demonstrated in real time by rapid magnitude estimation for the July 6, 2019 Mw7.1 Ridgecrest mainshock. Displacements from the Network of the Americas (NOTA) stations in the epicentral region were processed using precise point positioning with 1-2 s of latency and used to estimate a magnitude of Mw6.9±0.4 within 13 seconds of earthquake onset time, using peak ground displacement scaling relationships. GNSS instruments do not clip in the near field and have a signal to noise ratio that increases with the magnitude of displacement. Furthermore, they do not experience magnitude saturation. The precision of 2-3 centimeters for real-time (1 Hz) GNSS position estimates, however, is not sufficient to detect seismic P waves and issue warning of S wave or surface wave arrivals. By collocating strong motion accelerometers at GNSS stations and using a Kalman filter to estimate seismogeodetic velocities and displacements, we obtain a broadband record of the event from the Nyquist frequency of the accelerometers (typically 50 Hz) to the DC offset. We have collocated low-cost 12-bit MEMS accelerometers at 25 GNSS stations in southern California and the San Francisco Bay Area and successfully recorded P wave arrivals for four magnitude ~M4.5 earthquakes. In addition, we have demonstrated for the 2014 M8.2 Iquique subduction zone event off the coast of Chile that seismogeodetic velocities can be used to rapidly estimating magnitude using the Mwp method developed for traditional broadband seismometer data and implemented by NOAA’s Pacific Tsunami Warning Center. We also present the Mwp results for the 2019 Ridgecrest earthquake. Again, the advantage of seismogeodesy is that it provides a broadband instrument that does not clip, so that effective measurements are available closer to the earthquake’s source, thereby improving the timeliness of magnitude estimation.

Presenting Author: Yehuda Bock

Authors